[{"data":1,"prerenderedAt":818},["ShallowReactive",2],{"/en-us/blog/proposed-server-purchase-for-gitlab-com":3,"navigation-en-us":33,"banner-en-us":443,"footer-en-us":453,"blog-post-authors-en-us-Sid Sijbrandij":695,"blog-related-posts-en-us-proposed-server-purchase-for-gitlab-com":713,"blog-promotions-en-us":755,"next-steps-en-us":808},{"id":4,"title":5,"authorSlugs":6,"body":8,"categorySlug":9,"config":10,"content":14,"description":8,"extension":22,"isFeatured":12,"meta":23,"navigation":24,"path":25,"publishedDate":20,"seo":26,"stem":30,"tagSlugs":31,"__hash__":32},"blogPosts/en-us/blog/proposed-server-purchase-for-gitlab-com.yml","Proposed Server Purchase For Gitlab Com",[7],"sid-sijbrandij",null,"engineering",{"slug":11,"featured":12,"template":13},"proposed-server-purchase-for-gitlab-com",false,"BlogPost",{"title":15,"description":16,"authors":17,"heroImage":19,"date":20,"body":21,"category":9},"Proposed server purchase for GitLab.com","What hardware we're considering purchasing now that we have to move GitLab.com to metal.",[18],"Sid Sijbrandij","https://res.cloudinary.com/about-gitlab-com/image/upload/v1749666262/Blog/Hero%20Images/default-blog-image.png","2016-12-11","\n\nWe want to make GitLab.com fast and we [knew it was time to leave the cloud](/blog/why-choose-bare-metal/) and purchase our own servers.\nIn this post is our thinking about what chassis, rack, memory, CPU, network, power, and hosting to buy.\nWe wanted to share what we learned and get your feedback on our proposal and questions.\nWhen you reply to a question in the comments on our blog or Hacker News please reference it with the letter and number: 'Regarding R1'.\nWe'll try to update the questions with preliminary answers as we learn more.\n\n\u003C!-- more -->\n\n## Overview\n\nToday, GitLab.com hosts 96TB of data, and that number is growing rapidly. We\nare attempting to build a fault-tolerant and performant CephFS cluster. We are\nalso attempting to move GitLab application servers and supporting services\n(e.g. PostgreSQL) to bare metal.\n\nNote that for now our CI Runners will stay in the cloud. Not only are they are\nmuch less sensitive to latency, but autoscaling is easier with a cloud service.\n\n### Chassis\n\nOne of the team members that will join GitLab in 2017 recommended using a [6028TP-HTTR SuperMicro 2U Twin2 server](https://www.supermicro.nl/products/system/2U/6028/SYS-6028TP-HTTR.cfm) chassis that has 4 dual processor nodes and is 2 [rack units](https://en.wikipedia.org/wiki/Rack_unit) (U) high. The advantages are:\n\n1. Great density, 0.5U per dual processor server\n1. You have one common form factor\n1. Power supplies are shared for great efficiency similar to [blade servers](https://en.wikipedia.org/wiki/Blade_server)\n1. The network is per node for more bandwidth and reliability (like individual server)\n\nWe use the [2U Twin2](https://www.supermicro.com/products/nfo/2UTwin2.cfm) instead of the [1U Twin](https://www.supermicro.com/products/nfo/1UTwin.cfm) because it fits one more 3.5\" hard drive (3 per node instead of 2).\n\nThis server is on the list of global SKU's for SuperMicro.\nWe'll also ask for quotes from other vendors to see if they have a competitive alternative.\nFor example HPE has the [Apollo 2000 series](https://www.hpe.com/h20195/v2/getpdf.aspx/c04542552.pdf?ver=7).\n\nC1 Should we use another version of the chassis than HTTR?\n\nC2 What is the best Dell equivalent? => [C6320](http://www.dell.com/us/business/p/poweredge-c6320/pd)\n\n### Servers\n\nWe need the following servers:\n\n1. 32x File storage (CephFS OSD)\n1. 3x File Monitoring (CephFS MON)\n1. 8x Application server ([Unicorn](https://bogomips.org/unicorn/))\n1. 7x Background jobs ([Sidekiq](http://sidekiq.org/))\n1. 5x Key value store ([Redis Sentinel](https://redis.io/topics/sentinel))\n1. 4x Database (PostgreSQL)\n1. 3x Load balancers (HAproxy)\n1. 1x Staging\n1. 1x Spare\n\nFor a total of 64 nodes.\n\nWe would like to have one common node so that they are interchangeable.\nThis would mean installing only a few disks per node instead of having large fileservers.\nThis would distribute failures and IO.\n\n![IOPS on GitLab.com](https://about.gitlab.com/images/blogimages/write_iops.png)\n\nThe above picture shows the currently number of Input/output Operations Per\nSecond (IOPS) on GitLab.com. On our current NFS servers, our peak write IOPS\noften hit close to 500K, and our peak read IOPS reach 200K. These numbers\nsuggest that using spinning disks alone may not be enough; we need to use\nhigh-performance SSDs judiciously.\n\nOne task that we could not fit on the common nodes was PostgreSQL.\nOur current plan is to make PostgreSQL distributed in 2017 with the help of [Citus](https://www.citusdata.com/).\nBut for now, we need to scale vertically so we need a lot of memory and CPU.\nWe need at least a primary and secondary database.\nWe wanted to add a second pair for testing and to ensure spares in case of failure.\nDetails about this are in the following sections.\n\nChoosing a common node will mean that file storage servers will have too much CPU and that application servers will have too much disk space.\nWe plan to remedy that by running everything on Kubernetes.\nThis allows us to have a blended workload using all CPU and disk.\nFor example we can combine file storage and background jobs on the same server since one is disk heavy and one is CPU heavy.\nWe will start by having one workload per server to reduce complexity.\nThis means that when we need to grow we can still unlock almost twice as much disk space and CPU by blending the workloads.\nPlease note that this will be container based, to get maximum IO performance we won't virtualize our workload.\n\nS1 Shall we spread the database servers among different chassis to make sure they don't all fail when one chassis fails?\n\nS2 Does Ceph handle running 60 OSD nodes well or can this cause problems?\n\n### CPU\n\nThe [SuperServer 6028TP-HTTR](https://www.supermicro.nl/products/system/2U/6028/SYS-6028TP-HTTR.cfm) supports dual E5-2600v4 processors per node.\nWe think the [E5-2630v4](http://ark.intel.com/products/92981/Intel-Xeon-Processor-E5-2630-v4-25M-Cache-2_20-GHz) is a good blend of power and cost.\nIt has 20 virtual cores at 2.20Ghz, 25MB cache, and costs about $669 per processor.\nEvery physical core is two virtual cores due to [hyperthreading](https://en.wikipedia.org/wiki/Hyper-threading).\nA slightly more powerful processor is the [E5-2640v4](https://ark.intel.com/products/92984/Intel-Xeon-Processor-E5-2640-v4-25M-Cache-2_40-GHz) but while the [SPECint score](https://en.wikipedia.org/wiki/SPECint) increases from 845 to 887 the costs increase from $669 to $939.\nYou can find the scores by entering a [search on spec.org](https://www.spec.org/cgi-bin/osgresults?conf=rint2006) with 'Hewlett Packard Enterprise' as the hardware vendor and looking for ProLiant DL360 Gen9 as the platform.\n\nOur current SQL server has one E5-2698B v3 with 32 virtual cores.\nPostgreSQL commonly uses about 20-25 virtual cores.\nMoving to dual processors should already help a lot.\nTo give us more months to grow before having to distribute the database we want to purchase some headroom.\nThat is why we're getting a [E5-2687Wv4](https://ark.intel.com/products/91750/Intel-Xeon-Processor-E5-2687W-v4-30M-Cache-3_00-GHz) for the database servers.\nThis processor costs $2100 instead of $670 but has 4 extra virtual cores and runs continuously on 3 Ghz instead of 2.2 Ghz.\nComprated to the E5-2630v4 that leads to a SPEC score or 1230 instead of 845 and 51.3 SPEC per virtual core instead of 42.3.\nFor the 4 dual processor database servers this upgrade will cost $11k.\nWe think it is worth it since the 20-40% of extra performance will buy us the month or two of extra time to distribute the database that we need.\n\n### Disk\n\nEvery node can fit 3 larger (3.5\") harddrives.\nWe plan to purchase the largest one available, a 8TB Seagate with 6Gb/s SATA and 7.2K RPM.\nAt 60 nodes this will give us 1.4PB of raw storage.\nAt a replication factor of 3 for Ceph this is 480TB of usable storage.\nRight now GitLab.com uses 96TB (54TB for repo's, 21TB for uploads, 21TB for LFS and build artifacts) so we can grow by a factor of almost 5.\n\nDisks can be slow so we looked at improving latency.\nHigher RPM hard drives typically come in [GB instead of TB sizes](http://www.seagate.com/enterprise-storage/hard-disk-drives/enterprise-performance-15k-hdd/).\nGoing all SSD is too expensive.\nTo improve latency we plan to fit every server with an SSD card.\nOn the fileservers this will be used as a cache.\nWe're thinking about using [Bcache](https://en.wikipedia.org/wiki/Bcache) for this.\n\nWe plan to use [Intel DC P3700 series](http://www.intel.com/content/www/us/en/solid-state-drives/ssd-dc-p3700-spec.html) or slight less powerful [P3600 series](http://www.intel.com/content/www/us/en/solid-state-drives/ssd-dc-p3600-spec.html) of SSD's because they are recommended by the CephFS experts we hired.\nFor most servers it will be the [800GB SSDPEDMD800G4](http://www.supermicro.com/products/nfo/PCI-E_SSD.cfm?show=Intel).\nFor the database servers we plan to use the 1.6TB variant to have more headroom.\nThe endurance we need for the database server is 90TB/year, the 3600 series is already above 4PB of endurance.\n\nWe plan to add a 64GB [SSD SATADOM boot drive](https://www.supermicro.com/products/nfo/SATADOM.cfm) to the servers to boot from.\nThis way we can keep the large SSD as a separate volume.\n\nD1 We plan to configure the disks as just a bunch of disks (JBOD) but heard that this caused performance problems with some controllers. Is this likely to impact us?\n\nD2 Should we use Bcache to improve latency on the Ceph OSD servers with SSD? => Make sure you're using a kernel >= 4.5, since that's when a bunch of stability patches landed (https://lkml.org/lkml/2015/12/5/38).\n\nD3 We heard concerns about fitting the PCIe 3.0 x 4 SSD card into [our chassis](https://www.supermicro.nl/products/system/2U/6028/SYS-6028TP-HTTR.cfm) that supports a PCI-E 3.0 x16 Low-profile slot. Will this fit? => [Florian Heigl](http://disq.us/p/1eedj2n): \"Somewhat unlikely you will be able to fit a P3700. I have a Twin^2 too and the only SSD I could fit there was a consumer NVME with a PCIe adapter board.\"\n\nD4 Should we ask for 8TB HGST drives instead of Seagate since they seem [more reliable](https://www.backblaze.com/blog/hard-drive-reliability-stats-q1-2016/).\n\nD5 Is it a good idea to have a boot drive or should we use [PXE boot](https://en.wikipedia.org/wiki/Preboot_Execution_Environment) every time it starts? => [dsr_](https://news.ycombinator.com/item?id=13153336): You want a local boot drive, and you want it to fall back to PXE booting if the local drive is unavailable. Your PXE image should default to the last known working image, and have a boot-time menu with options for a rescue image and an installer for your distribution of choice.\n\nD6 Should we go for the 3700 series SSD or save some money and go for the 3600 series? Both for the normal and the SQL servers?\n\nD7 We're planning on one SSD per node. For the OSD nodes (file server) that would mean having the Ceph journal and bcache on the same SSD. Is this a good idea?\n\n### Memory\n\nSuppose one node runs both as application server and fileserver.\nWe recommend virtual cores + 1 instances of Unicorn of about 0.5GB each, for a total of 21GB per node (2 processors * 21 unicorns per processor * 0.5GB).\nCeph recommends about 1GB per TB of data which comes out to 24 per node.\nSo theoretically we can fit everything in 45GB so 64GB should be enough.\n\nBut in practice we've seen 24TB OSD nodes use 79GB of memory.\nAnd the rule of thumb is have about 2GB per virtual core for background jobs available (40GB).\nSo in order not to be to low we'll spend the extra $30k to have 128GB of ECC memory per node instead of 64GB.\n\nFor the SQL nodes we'll need much more memory, we currently give it 440GB and it uses all of that.\nThe database is about 250GB in size and growing with 40GB per month.\nAt 250GB of server memory we redlined the server, probably because it no longer fits into memory.\nTheoretically the server supports 2TB of memory but it needs to fit in 16 memory slots per node.\nWe wanted to start with 1TB per server but we're not sure if we should go from a 64GB DIMM to 128GB to be able to expand later.\nBy having only half of the memory banks full you get half the bandwidth.\nAnd 64GB DIMMs already cost twice as much per GB as 32GB DIMMs, let alone 128GB ones.\nAt a price of about $940 per 64 DIMM the cost for 1TB of memory already is $15k per server.\n\nNote that larger sizes such as 64GB come in the form of LRDIMM that has a [small performance penalty](https://www.microway.com/hpc-tech-tips/ddr4-rdimm-lrdimm-performance-comparison/) but this looks acceptable.\n\nM1. Should we use 128GB DIMMS to be able to expand the database server later even though the will double the cost and half the bandwidth?\n\n### Network\n\nThe servers come with 2x 10Gbps RJ45 by default (Intel X540 Dual port 10GBase-T).\nWe want to [dual bound](https://docs.oracle.com/cd/E37670_01/E41138/html/ch11s05.html) the network connections to increase performance and reliability.\nThis will allow us to take routers out of service during low traffic times, for example to restart them after a software upgrade.\nWe think that 20Gbps is enough bandwidth to handle our data access and replication needs, right now our highest peaks are 1 Gbps.\nThis is important because we want to have minimal latency between the Ceph servers so network congestion would be a problem.\n\nCeph reference designs recommend a separated front and back network with the back network reserved for Ceph traffic.\nWe think that this is not needed as long as there is enough capacity.\nWe do want to have user request termination in a DMZ, so our HA proxy servers will be the only ones with a public IP.\n\nEach of the two physical network connections will connect to a different top of rack router.\nWe want to get a Software Defined Networking (SDN) compatible router so we have flexibility there.\nWe're considering the [10/40GbE SDN SuperSwitch (SSE-X3648S/SSE-X3648SR)](https://www.supermicro.com/products/accessories/Networking/SSE-X3648S.cfm) that can switch 1440 Gbps.\n\nApart from those routers we'll have a separate router for a 1Gbps management network.\nFor example to make [STONITH](https://en.wikipedia.org/wiki/STONITH) reliable when there is a lot of traffic on the normal network.\nEach node already has a separate 1Gbps connection for this.\n\nWe have 64+1 nodes (1 for backup) and most routers seem to have 48 ports.\nEvery node has 2 network ports so that is a need for 130 ports in total.\nWe're not use if we can use 3 routers with 48 ports each (144 in total) to cover that.\n\nN1 Which router should we purchase?\n\nN2 How do we interconnect the routers while keeping the network simple and fast?\n\nN3 Should we have a separate network for Ceph traffic?\n\nN4 Do we need an SDN compatible router or can we purchase something more affordable?\n\nN5 What router should we use for the management network?\n\n### Backup\n\nWe're still early in figuring out the backup solution so there are still lots of questions.\n\nBacking up 480TB of data (expected size in 2017) is pretty hard.\nWe thought about using [Google Nearline](https://cloud.google.com/storage-nearline/) because with a price of $0.01 per GB per month means that for $4800 we don't have to worry about much.\nBut restoring that over a 1Gbps connection takes 44 days, way too long.\n\nWe mainly want our backup to protect us against human and software errors.\nBecause all the files are already replicated 3 times hardware errors are unlikely to affect us.\nOf course we should have a good [Ceph CRUSH map](http://docs.ceph.com/docs/jewel/rados/operations/crush-map/) to prevent storing multiple copies on the same chassis.\n\nWe're most afraid of human error or Ceph corruption. For that reason we don't want to replicate on the Ceph level but on the file level.\n\nWe're thinking about using [Bareos backup software](https://www.bareos.org/en/) to replicate to a huge fileserver.\nWe're inspired by the posts about the [latest 480TB Backblaze storage pod 6.0](https://www.backblaze.com/blog/open-source-data-storage-server/) and these are available for $6k without drives from [Backuppods](https://www.backuppods.com/).\nBut SuperMicro offers a [comparable solution in the form of a SuperChassis that can hold 90 drives](https://www.supermicro.com/products/chassis/4U/946/SC946ED-R2KJBOD).\nAt 8TB per drive that is 720TB of raw storage.\nEven with RAID overhead it should be possible to have 480TB of usable storage (66%).\n\nThe SuperChassis is only hard drives, it still needs a controller. In a [reference architecture by Nexenta (PDF download)](https://nexenta.com/sites/default/files/docs/Nexenta_SMC_RA_DataSheet.pdf) two [SYS6028U](https://www.supermicro.com/products/system/2u/6028/sys-6028u-tr4_.cfm) with E5-2643v3 processors and 256GB of RAM is recommended. Unlike smaller configurations this one doesn't come with an SSD for [ZFS L2ARC](https://blogs.oracle.com/brendan/entry/test).\n\nSince backups are mostly linear we don't need an SSD for caching. In general 1GB of memory per TB of raw ZFS disk space is recommended. That would mean getting 512GB of RAM, 16x 32GB. Unlike the reference architecture we'll go with one controller. We're considering the [SuperServer 1028R-WC1RT](https://www.supermicro.com/products/system/1U/1028/SYS-1028R-WC1RT.cfm) since it is similar to our other servers, 1U, has 2x 10Gbps, 16 DIMM slots, and has 2 PCI slots. We'll use our regular [E5-2630v4](http://ark.intel.com/products/92981/Intel-Xeon-Processor-E5-2630-v4-25M-Cache-2_20-GHz) processor.\n\nThe question is if this controller can saturate the 20 Gbps uplink.\nFor this it needs to use both 12 Gbps SAS buses.\nAnd each drive has to do at least 30 MBps which seems reasonable for a continuous read.\n\nThe problem is that even at 20Gbps a full restore takes 2 days.\nOf course many times you need to restore only part of the files (uploads).\nAnd most of the time it won't contain 480TB (we'll start at about 100TB).\nThe question is if we can accept this worst case scenario for GitLab.com.\n\nAn alternative would be to use multiple controllers.\nBut you can't aggregate ZFS pools over multiple servers.\nAnother option would be to have one controller with more IO.\nWe can use multiple disk enclosures and multiple SAS buses.\nAnd we can add more network ports and/or switch to 40Gbps.\nBut this all seems pretty complicated.\n\nB0 Are we on the right track here or is 20 Gbps of restore speed not OK?\n\nB1 Should we go for the [90 or 60 drive SuperChassis](https://www.supermicro.com/products/chassis/4U/?chs=946)? It looks like 60 drive one has more peak power (1600W vs. 800W) to start the drives.\n\nB2 How should we configure the SuperChassis? [ZFS on Linux](http://zfsonlinux.org/) with [RAIDZ3](https://icesquare.com/wordpress/zfs-performance-mirror-vs-raidz-vs-raidz2-vs-raidz3-vs-striped/)?\n\nB3 Will the SuperChassis be able to saturate the 20Gbsp connection?\n\nB4 Should we upgrade the networking on the SuperChassis to be able to restore even faster?\n\nB5 Is Bareos the right software to use?\n\nB6 How should we configure the backup software?  Should we use incremental backups with parallel jobs to speed things up?\n\nB7 Should we use the live filesystem or [CephFS snapshots](http://docs.ceph.com/docs/master/dev/cephfs-snapshots/) to back up from?\n\nB8 How common is it to have a tape or cloud backup in addition to the above?\n\nB9 Should we pick the top load model or [one of the front and rear access models](https://www.supermicro.com/products/chassis/JBOD/index.cfm?show=SELECT&storage=90).\n\nB10 Can we connect two SAS cables to get 2x 12 Gbps?\n\nB11 What [HBA card](https://www.supermicro.com/products/nfo/storage_cards.cfm) should be added to the controller or does it come with an LSI 3108?\n\nB12 Is it smart to make the controller a separate 1U box or should we repurpose some of our normal nodes for this?\n\nB13 Any hints on how to test the backup restore (on AWS or our hardware, how often, etc.)?\n\n### Rack\n\nThe default rack height seems to be 45U nowadays (42U used to be the standard).\n\nIt is used as follows:\n\n- 32U for 16 chassis with 64 nodes\n- 3U for three network routers\n- 1U for the management network\n- 4U for the disk enclosure\n- 1U for the disk controller\n- 4U spare for 2 new chassis (maybe distributed PostgreSQL servers)\n\n### Power\n\nEach chassis has a 2000 watt power supply (comes to 1kW per U), 32kW in total.\nNormal usage is guessed at 60% of the rated capacity, about 19kW.\nThat doesn't account for the routers and backup.\nBoth hosting providers quoted 4 x 208v 30A power supplies (2 for redundancy).\n\nP1 Does the quoted supply seem adequate for our needs?\n\n### Hosting\n\nWe've worked in [an issue](https://gitlab.com/gitlab-com/infrastructure/issues/732) to see where we should host.\n\nApart from the obvious (reliable, affordable) we had the following needs:\n\n- [AWS Direct connect](https://aws.amazon.com/directconnect/details/) so we can use the cloud for temporary application server needs\n- Based on the east coast of the USA since it provides the best latency tradeoff for most of our users\n- Advanced remote hands service so we don't have to station people near the datacenter at all times\n- Ability to upgrade from one rack to a private cage\n\nThe following networking options are a plus:\n\n- Carrier neutral (all major global network providers in its meet-me facility)\n- Backbones to other locations to provide cheap 2nd site transit\n- CDN services to reduce origin bandwidth costs\n\nSo far we've gotten quotes from [QTS in Ashburn, VA](http://www.qtsdatacenters.com/data-centers/ashburn) and [NYI in Bridgewater, NJ](https://www.nyi.net/datacenters/new-jersey/).\n\nH1 Any watchouts when selecting hosting providers?\n\nH2 Should we install the servers ourselves or is it OK to let the hosting provider do that?\n\nH3 How can we minimize installation costs? Should we ask to configure the servers to PXE boot?\n\nH4 Is there an Azure equivalent for AWS Direct Connect? => Azure will let you work with a provider to \"peer into\" the Azure network at a data center of your choice. So for example we could pay to have a circuit established in a data center that was linked into the Azure 'US East 2' data center (where we currently host out of) for direct connectivity needs.\n\n### Expense\n\nWe can't give cost details since all the quotes we receive are confidential.\nThe cloud hosting for GitLab.com excluding GitLab CI is currently costing us about $200k per month.\nThe capital needed for going to metal would be less than we pay for 1 quarter of hosting.\nThe hosting facility costs look to be less than $10k per month.\nIf you spread the capital costs over 2.5 years (10 quarters) it is 10x cheaper to host your own.\n\nOf course the growth of GitLab.com will soon force us to buy additional hardware.\nBut we would also have to pay extra for additional cloud capacity.\nOur proposed buying plan is about 5x the capacity we need now.\nHaving your own hardware means you're always overprovisioned.\nAnd we could probably have reduced the cost of cloud hosting by focussing on it.\n\nThe bigger expense will be hiring more people to deal with the additional complexity.\nWe'll probably need to hire a couple of people more to deal with this.\n\nWe looked into initially having disks in only half the servers but that saves only $20k ($225 per disk) and it would create a lot of work when we eventually have to install them.\n\nE1 If we want to look at leasing should we do that through SuperMicro or third party?\n\nE2 Are there ways we can save money?\n\n## Details\n\nOur detailed calculations and notes can be found in a [public Google sheet](https://docs.google.com/spreadsheets/d/1XG9VXdDxNd8ipgPlEr7Nb7Eg22twXPuzgDwsOhtdYKQ/edit#gid=894825456).\n","yml",{},true,"/en-us/blog/proposed-server-purchase-for-gitlab-com",{"title":15,"description":16,"ogTitle":15,"ogDescription":16,"noIndex":12,"ogImage":19,"ogUrl":27,"ogSiteName":28,"ogType":29,"canonicalUrls":27},"https://about.gitlab.com/blog/proposed-server-purchase-for-gitlab-com","https://about.gitlab.com","article","en-us/blog/proposed-server-purchase-for-gitlab-com",[],"3jaZ_8SIungJiTffc0cgB18LKbgYY0UToJz4_ztR_X0",{"data":34},{"logo":35,"freeTrial":40,"sales":45,"login":50,"items":55,"search":363,"minimal":394,"duo":413,"switchNav":422,"pricingDeployment":433},{"config":36},{"href":37,"dataGaName":38,"dataGaLocation":39},"/","gitlab logo","header",{"text":41,"config":42},"Get free trial",{"href":43,"dataGaName":44,"dataGaLocation":39},"https://gitlab.com/-/trial_registrations/new?glm_source=about.gitlab.com&glm_content=default-saas-trial/","free trial",{"text":46,"config":47},"Talk to sales",{"href":48,"dataGaName":49,"dataGaLocation":39},"/sales/","sales",{"text":51,"config":52},"Sign in",{"href":53,"dataGaName":54,"dataGaLocation":39},"https://gitlab.com/users/sign_in/","sign in",[56,83,178,183,284,344],{"text":57,"config":58,"cards":60},"Platform",{"dataNavLevelOne":59},"platform",[61,67,75],{"title":57,"description":62,"link":63},"The intelligent orchestration platform for DevSecOps",{"text":64,"config":65},"Explore our Platform",{"href":66,"dataGaName":59,"dataGaLocation":39},"/platform/",{"title":68,"description":69,"link":70},"GitLab Duo Agent Platform","Agentic AI for the entire software lifecycle",{"text":71,"config":72},"Meet GitLab Duo",{"href":73,"dataGaName":74,"dataGaLocation":39},"/gitlab-duo-agent-platform/","gitlab duo agent platform",{"title":76,"description":77,"link":78},"Why GitLab","See the top reasons enterprises choose GitLab",{"text":79,"config":80},"Learn more",{"href":81,"dataGaName":82,"dataGaLocation":39},"/why-gitlab/","why gitlab",{"text":84,"left":24,"config":85,"link":87,"lists":91,"footer":160},"Product",{"dataNavLevelOne":86},"solutions",{"text":88,"config":89},"View all Solutions",{"href":90,"dataGaName":86,"dataGaLocation":39},"/solutions/",[92,116,139],{"title":93,"description":94,"link":95,"items":100},"Automation","CI/CD and automation to accelerate deployment",{"config":96},{"icon":97,"href":98,"dataGaName":99,"dataGaLocation":39},"AutomatedCodeAlt","/solutions/delivery-automation/","automated software delivery",[101,105,108,112],{"text":102,"config":103},"CI/CD",{"href":104,"dataGaLocation":39,"dataGaName":102},"/solutions/continuous-integration/",{"text":68,"config":106},{"href":73,"dataGaLocation":39,"dataGaName":107},"gitlab duo agent platform - product menu",{"text":109,"config":110},"Source Code Management",{"href":111,"dataGaLocation":39,"dataGaName":109},"/solutions/source-code-management/",{"text":113,"config":114},"Automated Software Delivery",{"href":98,"dataGaLocation":39,"dataGaName":115},"Automated software delivery",{"title":117,"description":118,"link":119,"items":124},"Security","Deliver code faster without compromising security",{"config":120},{"href":121,"dataGaName":122,"dataGaLocation":39,"icon":123},"/solutions/application-security-testing/","security and compliance","ShieldCheckLight",[125,129,134],{"text":126,"config":127},"Application Security Testing",{"href":121,"dataGaName":128,"dataGaLocation":39},"Application security testing",{"text":130,"config":131},"Software Supply Chain Security",{"href":132,"dataGaLocation":39,"dataGaName":133},"/solutions/supply-chain/","Software supply chain security",{"text":135,"config":136},"Software Compliance",{"href":137,"dataGaName":138,"dataGaLocation":39},"/solutions/software-compliance/","software compliance",{"title":140,"link":141,"items":146},"Measurement",{"config":142},{"icon":143,"href":144,"dataGaName":145,"dataGaLocation":39},"DigitalTransformation","/solutions/visibility-measurement/","visibility and measurement",[147,151,155],{"text":148,"config":149},"Visibility & Measurement",{"href":144,"dataGaLocation":39,"dataGaName":150},"Visibility and Measurement",{"text":152,"config":153},"Value Stream Management",{"href":154,"dataGaLocation":39,"dataGaName":152},"/solutions/value-stream-management/",{"text":156,"config":157},"Analytics & Insights",{"href":158,"dataGaLocation":39,"dataGaName":159},"/solutions/analytics-and-insights/","Analytics and insights",{"title":161,"items":162},"GitLab for",[163,168,173],{"text":164,"config":165},"Enterprise",{"href":166,"dataGaLocation":39,"dataGaName":167},"/enterprise/","enterprise",{"text":169,"config":170},"Small Business",{"href":171,"dataGaLocation":39,"dataGaName":172},"/small-business/","small business",{"text":174,"config":175},"Public Sector",{"href":176,"dataGaLocation":39,"dataGaName":177},"/solutions/public-sector/","public sector",{"text":179,"config":180},"Pricing",{"href":181,"dataGaName":182,"dataGaLocation":39,"dataNavLevelOne":182},"/pricing/","pricing",{"text":184,"config":185,"link":187,"lists":191,"feature":271},"Resources",{"dataNavLevelOne":186},"resources",{"text":188,"config":189},"View all resources",{"href":190,"dataGaName":186,"dataGaLocation":39},"/resources/",[192,225,243],{"title":193,"items":194},"Getting started",[195,200,205,210,215,220],{"text":196,"config":197},"Install",{"href":198,"dataGaName":199,"dataGaLocation":39},"/install/","install",{"text":201,"config":202},"Quick start guides",{"href":203,"dataGaName":204,"dataGaLocation":39},"/get-started/","quick setup checklists",{"text":206,"config":207},"Learn",{"href":208,"dataGaLocation":39,"dataGaName":209},"https://university.gitlab.com/","learn",{"text":211,"config":212},"Product documentation",{"href":213,"dataGaName":214,"dataGaLocation":39},"https://docs.gitlab.com/","product documentation",{"text":216,"config":217},"Best practice videos",{"href":218,"dataGaName":219,"dataGaLocation":39},"/getting-started-videos/","best practice videos",{"text":221,"config":222},"Integrations",{"href":223,"dataGaName":224,"dataGaLocation":39},"/integrations/","integrations",{"title":226,"items":227},"Discover",[228,233,238],{"text":229,"config":230},"Customer success stories",{"href":231,"dataGaName":232,"dataGaLocation":39},"/customers/","customer success stories",{"text":234,"config":235},"Blog",{"href":236,"dataGaName":237,"dataGaLocation":39},"/blog/","blog",{"text":239,"config":240},"Remote",{"href":241,"dataGaName":242,"dataGaLocation":39},"https://handbook.gitlab.com/handbook/company/culture/all-remote/","remote",{"title":244,"items":245},"Connect",[246,251,256,261,266],{"text":247,"config":248},"GitLab Services",{"href":249,"dataGaName":250,"dataGaLocation":39},"/services/","services",{"text":252,"config":253},"Community",{"href":254,"dataGaName":255,"dataGaLocation":39},"/community/","community",{"text":257,"config":258},"Forum",{"href":259,"dataGaName":260,"dataGaLocation":39},"https://forum.gitlab.com/","forum",{"text":262,"config":263},"Events",{"href":264,"dataGaName":265,"dataGaLocation":39},"/events/","events",{"text":267,"config":268},"Partners",{"href":269,"dataGaName":270,"dataGaLocation":39},"/partners/","partners",{"backgroundColor":272,"textColor":273,"text":274,"image":275,"link":279},"#2f2a6b","#fff","Insights for the future of software development",{"altText":276,"config":277},"the source promo card",{"src":278},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1758208064/dzl0dbift9xdizyelkk4.svg",{"text":280,"config":281},"Read the latest",{"href":282,"dataGaName":283,"dataGaLocation":39},"/the-source/","the source",{"text":285,"config":286,"lists":288},"Company",{"dataNavLevelOne":287},"company",[289],{"items":290},[291,296,302,304,309,314,319,324,329,334,339],{"text":292,"config":293},"About",{"href":294,"dataGaName":295,"dataGaLocation":39},"/company/","about",{"text":297,"config":298,"footerGa":301},"Jobs",{"href":299,"dataGaName":300,"dataGaLocation":39},"/jobs/","jobs",{"dataGaName":300},{"text":262,"config":303},{"href":264,"dataGaName":265,"dataGaLocation":39},{"text":305,"config":306},"Leadership",{"href":307,"dataGaName":308,"dataGaLocation":39},"/company/team/e-group/","leadership",{"text":310,"config":311},"Team",{"href":312,"dataGaName":313,"dataGaLocation":39},"/company/team/","team",{"text":315,"config":316},"Handbook",{"href":317,"dataGaName":318,"dataGaLocation":39},"https://handbook.gitlab.com/","handbook",{"text":320,"config":321},"Investor relations",{"href":322,"dataGaName":323,"dataGaLocation":39},"https://ir.gitlab.com/","investor relations",{"text":325,"config":326},"Trust Center",{"href":327,"dataGaName":328,"dataGaLocation":39},"/security/","trust center",{"text":330,"config":331},"AI Transparency Center",{"href":332,"dataGaName":333,"dataGaLocation":39},"/ai-transparency-center/","ai transparency center",{"text":335,"config":336},"Newsletter",{"href":337,"dataGaName":338,"dataGaLocation":39},"/company/contact/#contact-forms","newsletter",{"text":340,"config":341},"Press",{"href":342,"dataGaName":343,"dataGaLocation":39},"/press/","press",{"text":345,"config":346,"lists":347},"Contact us",{"dataNavLevelOne":287},[348],{"items":349},[350,353,358],{"text":46,"config":351},{"href":48,"dataGaName":352,"dataGaLocation":39},"talk to sales",{"text":354,"config":355},"Support portal",{"href":356,"dataGaName":357,"dataGaLocation":39},"https://support.gitlab.com","support portal",{"text":359,"config":360},"Customer portal",{"href":361,"dataGaName":362,"dataGaLocation":39},"https://customers.gitlab.com/customers/sign_in/","customer portal",{"close":364,"login":365,"suggestions":372},"Close",{"text":366,"link":367},"To search repositories and projects, login to",{"text":368,"config":369},"gitlab.com",{"href":53,"dataGaName":370,"dataGaLocation":371},"search login","search",{"text":373,"default":374},"Suggestions",[375,377,381,383,387,391],{"text":68,"config":376},{"href":73,"dataGaName":68,"dataGaLocation":371},{"text":378,"config":379},"Code Suggestions (AI)",{"href":380,"dataGaName":378,"dataGaLocation":371},"/solutions/code-suggestions/",{"text":102,"config":382},{"href":104,"dataGaName":102,"dataGaLocation":371},{"text":384,"config":385},"GitLab on AWS",{"href":386,"dataGaName":384,"dataGaLocation":371},"/partners/technology-partners/aws/",{"text":388,"config":389},"GitLab on Google Cloud",{"href":390,"dataGaName":388,"dataGaLocation":371},"/partners/technology-partners/google-cloud-platform/",{"text":392,"config":393},"Why GitLab?",{"href":81,"dataGaName":392,"dataGaLocation":371},{"freeTrial":395,"mobileIcon":400,"desktopIcon":405,"secondaryButton":408},{"text":396,"config":397},"Start free trial",{"href":398,"dataGaName":44,"dataGaLocation":399},"https://gitlab.com/-/trials/new/","nav",{"altText":401,"config":402},"Gitlab Icon",{"src":403,"dataGaName":404,"dataGaLocation":399},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1758203874/jypbw1jx72aexsoohd7x.svg","gitlab icon",{"altText":401,"config":406},{"src":407,"dataGaName":404,"dataGaLocation":399},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1758203875/gs4c8p8opsgvflgkswz9.svg",{"text":409,"config":410},"Get Started",{"href":411,"dataGaName":412,"dataGaLocation":399},"https://gitlab.com/-/trial_registrations/new?glm_source=about.gitlab.com/get-started/","get started",{"freeTrial":414,"mobileIcon":418,"desktopIcon":420},{"text":415,"config":416},"Learn more about GitLab Duo",{"href":73,"dataGaName":417,"dataGaLocation":399},"gitlab duo",{"altText":401,"config":419},{"src":403,"dataGaName":404,"dataGaLocation":399},{"altText":401,"config":421},{"src":407,"dataGaName":404,"dataGaLocation":399},{"button":423,"mobileIcon":428,"desktopIcon":430},{"text":424,"config":425},"/switch",{"href":426,"dataGaName":427,"dataGaLocation":399},"#contact","switch",{"altText":401,"config":429},{"src":403,"dataGaName":404,"dataGaLocation":399},{"altText":401,"config":431},{"src":432,"dataGaName":404,"dataGaLocation":399},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1773335277/ohhpiuoxoldryzrnhfrh.png",{"freeTrial":434,"mobileIcon":439,"desktopIcon":441},{"text":435,"config":436},"Back to pricing",{"href":181,"dataGaName":437,"dataGaLocation":399,"icon":438},"back to pricing","GoBack",{"altText":401,"config":440},{"src":403,"dataGaName":404,"dataGaLocation":399},{"altText":401,"config":442},{"src":407,"dataGaName":404,"dataGaLocation":399},{"title":444,"button":445,"config":450},"See how agentic AI transforms software delivery",{"text":446,"config":447},"Watch GitLab Transcend now",{"href":448,"dataGaName":449,"dataGaLocation":39},"/events/transcend/virtual/","transcend event",{"layout":451,"icon":452,"disabled":24},"release","AiStar",{"data":454},{"text":455,"source":456,"edit":462,"contribute":467,"config":472,"items":477,"minimal":684},"Git is a trademark of Software Freedom Conservancy and our use of 'GitLab' is under license",{"text":457,"config":458},"View page source",{"href":459,"dataGaName":460,"dataGaLocation":461},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/","page source","footer",{"text":463,"config":464},"Edit this page",{"href":465,"dataGaName":466,"dataGaLocation":461},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/-/blob/main/content/","web ide",{"text":468,"config":469},"Please contribute",{"href":470,"dataGaName":471,"dataGaLocation":461},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/-/blob/main/CONTRIBUTING.md/","please contribute",{"twitter":473,"facebook":474,"youtube":475,"linkedin":476},"https://twitter.com/gitlab","https://www.facebook.com/gitlab","https://www.youtube.com/channel/UCnMGQ8QHMAnVIsI3xJrihhg","https://www.linkedin.com/company/gitlab-com",[478,525,579,623,650],{"title":179,"links":479,"subMenu":494},[480,484,489],{"text":481,"config":482},"View plans",{"href":181,"dataGaName":483,"dataGaLocation":461},"view plans",{"text":485,"config":486},"Why Premium?",{"href":487,"dataGaName":488,"dataGaLocation":461},"/pricing/premium/","why premium",{"text":490,"config":491},"Why Ultimate?",{"href":492,"dataGaName":493,"dataGaLocation":461},"/pricing/ultimate/","why ultimate",[495],{"title":496,"links":497},"Contact Us",[498,501,503,505,510,515,520],{"text":499,"config":500},"Contact sales",{"href":48,"dataGaName":49,"dataGaLocation":461},{"text":354,"config":502},{"href":356,"dataGaName":357,"dataGaLocation":461},{"text":359,"config":504},{"href":361,"dataGaName":362,"dataGaLocation":461},{"text":506,"config":507},"Status",{"href":508,"dataGaName":509,"dataGaLocation":461},"https://status.gitlab.com/","status",{"text":511,"config":512},"Terms of use",{"href":513,"dataGaName":514,"dataGaLocation":461},"/terms/","terms of use",{"text":516,"config":517},"Privacy statement",{"href":518,"dataGaName":519,"dataGaLocation":461},"/privacy/","privacy statement",{"text":521,"config":522},"Cookie preferences",{"dataGaName":523,"dataGaLocation":461,"id":524,"isOneTrustButton":24},"cookie preferences","ot-sdk-btn",{"title":84,"links":526,"subMenu":535},[527,531],{"text":528,"config":529},"DevSecOps platform",{"href":66,"dataGaName":530,"dataGaLocation":461},"devsecops platform",{"text":532,"config":533},"AI-Assisted Development",{"href":73,"dataGaName":534,"dataGaLocation":461},"ai-assisted development",[536],{"title":537,"links":538},"Topics",[539,544,549,554,559,564,569,574],{"text":540,"config":541},"CICD",{"href":542,"dataGaName":543,"dataGaLocation":461},"/topics/ci-cd/","cicd",{"text":545,"config":546},"GitOps",{"href":547,"dataGaName":548,"dataGaLocation":461},"/topics/gitops/","gitops",{"text":550,"config":551},"DevOps",{"href":552,"dataGaName":553,"dataGaLocation":461},"/topics/devops/","devops",{"text":555,"config":556},"Version Control",{"href":557,"dataGaName":558,"dataGaLocation":461},"/topics/version-control/","version control",{"text":560,"config":561},"DevSecOps",{"href":562,"dataGaName":563,"dataGaLocation":461},"/topics/devsecops/","devsecops",{"text":565,"config":566},"Cloud Native",{"href":567,"dataGaName":568,"dataGaLocation":461},"/topics/cloud-native/","cloud native",{"text":570,"config":571},"AI for Coding",{"href":572,"dataGaName":573,"dataGaLocation":461},"/topics/devops/ai-for-coding/","ai for coding",{"text":575,"config":576},"Agentic AI",{"href":577,"dataGaName":578,"dataGaLocation":461},"/topics/agentic-ai/","agentic ai",{"title":580,"links":581},"Solutions",[582,584,586,591,595,598,602,605,607,610,613,618],{"text":126,"config":583},{"href":121,"dataGaName":126,"dataGaLocation":461},{"text":115,"config":585},{"href":98,"dataGaName":99,"dataGaLocation":461},{"text":587,"config":588},"Agile development",{"href":589,"dataGaName":590,"dataGaLocation":461},"/solutions/agile-delivery/","agile delivery",{"text":592,"config":593},"SCM",{"href":111,"dataGaName":594,"dataGaLocation":461},"source code management",{"text":540,"config":596},{"href":104,"dataGaName":597,"dataGaLocation":461},"continuous integration & delivery",{"text":599,"config":600},"Value stream management",{"href":154,"dataGaName":601,"dataGaLocation":461},"value stream management",{"text":545,"config":603},{"href":604,"dataGaName":548,"dataGaLocation":461},"/solutions/gitops/",{"text":164,"config":606},{"href":166,"dataGaName":167,"dataGaLocation":461},{"text":608,"config":609},"Small business",{"href":171,"dataGaName":172,"dataGaLocation":461},{"text":611,"config":612},"Public sector",{"href":176,"dataGaName":177,"dataGaLocation":461},{"text":614,"config":615},"Education",{"href":616,"dataGaName":617,"dataGaLocation":461},"/solutions/education/","education",{"text":619,"config":620},"Financial services",{"href":621,"dataGaName":622,"dataGaLocation":461},"/solutions/finance/","financial services",{"title":184,"links":624},[625,627,629,631,634,636,638,640,642,644,646,648],{"text":196,"config":626},{"href":198,"dataGaName":199,"dataGaLocation":461},{"text":201,"config":628},{"href":203,"dataGaName":204,"dataGaLocation":461},{"text":206,"config":630},{"href":208,"dataGaName":209,"dataGaLocation":461},{"text":211,"config":632},{"href":213,"dataGaName":633,"dataGaLocation":461},"docs",{"text":234,"config":635},{"href":236,"dataGaName":237,"dataGaLocation":461},{"text":229,"config":637},{"href":231,"dataGaName":232,"dataGaLocation":461},{"text":239,"config":639},{"href":241,"dataGaName":242,"dataGaLocation":461},{"text":247,"config":641},{"href":249,"dataGaName":250,"dataGaLocation":461},{"text":252,"config":643},{"href":254,"dataGaName":255,"dataGaLocation":461},{"text":257,"config":645},{"href":259,"dataGaName":260,"dataGaLocation":461},{"text":262,"config":647},{"href":264,"dataGaName":265,"dataGaLocation":461},{"text":267,"config":649},{"href":269,"dataGaName":270,"dataGaLocation":461},{"title":285,"links":651},[652,654,656,658,660,662,664,668,673,675,677,679],{"text":292,"config":653},{"href":294,"dataGaName":287,"dataGaLocation":461},{"text":297,"config":655},{"href":299,"dataGaName":300,"dataGaLocation":461},{"text":305,"config":657},{"href":307,"dataGaName":308,"dataGaLocation":461},{"text":310,"config":659},{"href":312,"dataGaName":313,"dataGaLocation":461},{"text":315,"config":661},{"href":317,"dataGaName":318,"dataGaLocation":461},{"text":320,"config":663},{"href":322,"dataGaName":323,"dataGaLocation":461},{"text":665,"config":666},"Sustainability",{"href":667,"dataGaName":665,"dataGaLocation":461},"/sustainability/",{"text":669,"config":670},"Diversity, inclusion and belonging (DIB)",{"href":671,"dataGaName":672,"dataGaLocation":461},"/diversity-inclusion-belonging/","Diversity, inclusion and belonging",{"text":325,"config":674},{"href":327,"dataGaName":328,"dataGaLocation":461},{"text":335,"config":676},{"href":337,"dataGaName":338,"dataGaLocation":461},{"text":340,"config":678},{"href":342,"dataGaName":343,"dataGaLocation":461},{"text":680,"config":681},"Modern Slavery Transparency Statement",{"href":682,"dataGaName":683,"dataGaLocation":461},"https://handbook.gitlab.com/handbook/legal/modern-slavery-act-transparency-statement/","modern slavery transparency statement",{"items":685},[686,689,692],{"text":687,"config":688},"Terms",{"href":513,"dataGaName":514,"dataGaLocation":461},{"text":690,"config":691},"Cookies",{"dataGaName":523,"dataGaLocation":461,"id":524,"isOneTrustButton":24},{"text":693,"config":694},"Privacy",{"href":518,"dataGaName":519,"dataGaLocation":461},[696],{"id":697,"title":18,"body":8,"config":698,"content":700,"description":8,"extension":22,"meta":708,"navigation":24,"path":709,"seo":710,"stem":711,"__hash__":712},"blogAuthors/en-us/blog/authors/sid-sijbrandij.yml",{"template":699},"BlogAuthor",{"role":701,"name":18,"bio":702,"config":703},"Co-founder, Chief Executive Officer and Board Chair of GitLab Inc.","Sid Sijbrandij (pronounced see-brandy) is the Co-founder, Chief Executive Officer and Board Chair of GitLab Inc., the most comprehensive AI-powered DevSecOps platform. GitLab's single application helps organizations deliver software faster and more efficiently while strengthening their security and compliance.\n\nSid's career path has been anything but traditional. He spent four years building recreational submarines for U-Boat Worx and while at Ministerie van Justitie en Veiligheid he worked on the Legis project, which developed several innovative web applications to aid lawmaking. He first saw Ruby code in 2007 and loved it so much that he taught himself how to program. In 2012, as a Ruby programmer, he encountered GitLab and discovered his passion for open source. Soon after, Sid commercialized GitLab, and by 2015 he led the company through Y Combinator's Winter 2015 batch. Under his leadership, the company has grown with an estimated 30 million+ registered users from startups to global enterprises.\n\nSid studied at the University of Twente in the Netherlands where he received an M.S. in Management Science. Sid was named one of the greatest minds of the pandemic by Forbes for spreading the gospel of remote work.",{"headshot":704,"twitter":705,"linkedin":706,"ctfId":707},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1749665383/Blog/Author%20Headshots/sytses-headshot.png","https://twitter.com/sytses","https://www.linkedin.com/in/sijbrandij","sytses",{},"/en-us/blog/authors/sid-sijbrandij",{},"en-us/blog/authors/sid-sijbrandij","ZdVvFbtL6NKLtKZEjFCVOecdpvuPzX3wmEZBrC6pRWg",[714,729,742],{"content":715,"config":727},{"body":716,"title":717,"description":718,"authors":719,"heroImage":721,"date":722,"category":9,"tags":723},"Most CI/CD tools can run a build and ship a deployment. Where they diverge is what happens when your delivery needs get real: a monorepo with a dozen services, microservices spread across multiple repositories, deployments to dozens of environments, or a platform team trying to enforce standards without becoming a bottleneck.\n  \nGitLab's pipeline execution model was designed for that complexity. Parent-child pipelines, DAG execution, dynamic pipeline generation, multi-project triggers, merge request pipelines with merged results, and CI/CD Components each solve a distinct class of problems. Because they compose, understanding the full model unlocks something more than a faster pipeline. In this article, you'll learn about the five patterns where that model stands out, each mapped to a real engineering scenario with the configuration to match.\n  \nThe configs below are illustrative. The scripts use echo commands to keep the signal-to-noise ratio low. Swap them out for your actual build, test, and deploy steps and they are ready to use.\n\n\n## 1. Monorepos: Parent-child pipelines + DAG execution\n\n\nThe problem: Your monorepo has a frontend, a backend, and a docs site. Every commit triggers a full rebuild of everything, even when only a README changed.\n\n\nGitLab solves this with two complementary features: [parent-child pipelines](https://docs.gitlab.com/ci/pipelines/downstream_pipelines/#parent-child-pipelines) (which let a top-level pipeline spawn isolated sub-pipelines) and [DAG execution via `needs`](https://docs.gitlab.com/ci/yaml/#needs) (which breaks rigid stage-by-stage ordering and lets jobs start the moment their dependencies finish).\n\n\nA parent pipeline detects what changed and triggers only the relevant child pipelines:\n\n```yaml\n# .gitlab-ci.yml\nstages:\n  - trigger\n\ntrigger-services:\n  stage: trigger\n  trigger:\n    include:\n      - local: '.gitlab/ci/api-service.yml'\n      - local: '.gitlab/ci/web-service.yml'\n      - local: '.gitlab/ci/worker-service.yml'\n    strategy: depend\n```\n\n\nEach child pipeline is a fully independent pipeline with its own stages, jobs, and artifacts. The parent waits for all of them via [strategy: depend](https://docs.gitlab.com/ci/pipelines/downstream_pipelines/#wait-for-downstream-pipeline-to-complete) so you get a single green/red signal at the top level, with full drill-down into each service's pipeline. This organizational separation is the bigger win for large teams: each service owns its pipeline config, changes in one cannot break another, and the complexity stays manageable as the repo grows.\n\n\nOne thing worth knowing: when you pass [multiple files to a single `trigger: include:`](https://docs.gitlab.com/ci/pipelines/downstream_pipelines/#combine-multiple-child-pipeline-configuration-files), GitLab merges them into a single child pipeline configuration. This means jobs defined across those files share the same pipeline context and can reference each other with `needs:`, which is what makes the DAG optimization possible. If you split them into separate trigger jobs instead, each would be its own isolated pipeline and cross-file `needs:` references would not work.\n\n\nCombine this with `needs:` inside each child pipeline and you get DAG execution. Your integration tests can start the moment the build finishes, without waiting for other jobs in the same stage.\n\n```yaml\n# .gitlab/ci/api-service.yml\nstages:\n  - build\n  - test\n\nbuild-api:\n  stage: build\n  script:\n    - echo \"Building API service\"\n\ntest-api:\n  stage: test\n  needs: [build-api]\n  script:\n    - echo \"Running API tests\"\n```\n\n\nWhy it matters: Teams with large monorepos typically report significant reductions in pipeline runtime after switching to DAG execution, since jobs no longer wait on unrelated work in the same stage. Parent-child pipelines add the organizational layer that keeps the configuration maintainable as the repo and team grow.\n\n![Local downstream pipelines](https://res.cloudinary.com/about-gitlab-com/image/upload/v1775738759/Blog/Imported/hackathon-fake-blog-post-s/image3_vwj3rz.png \"Local downstream pipelines\")\n\n## 2. Microservices: Cross-repo, multi-project pipelines\n\n\nThe problem: Your frontend lives in one repo, your backend in another. When the frontend team ships a change, they have no visibility into whether it broke the backend integration and vice versa.\n\n\nGitLab's [multi-project pipelines](https://docs.gitlab.com/ci/pipelines/downstream_pipelines/#multi-project-pipelines) let one project trigger a pipeline in a completely separate project and wait for the result. The triggering project gets a linked downstream pipeline right in its own pipeline view.\n\n\nThe frontend pipeline builds an API contract artifact and publishes it, then triggers the backend pipeline. The backend fetches that artifact directly using the [Jobs API](https://docs.gitlab.com/ee/api/jobs.html#download-a-single-artifact-file-from-specific-tag-or-branch) and validates it before allowing anything to proceed. If a breaking change is detected, the backend pipeline fails and the frontend pipeline fails with it.\n\n```yaml\n# frontend repo: .gitlab-ci.yml\nstages:\n  - build\n  - test\n  - trigger-backend\n\nbuild-frontend:\n  stage: build\n  script:\n    - echo \"Building frontend and generating API contract...\"\n    - mkdir -p dist\n    - |\n      echo '{\n        \"api_version\": \"v2\",\n        \"breaking_changes\": false\n      }' > dist/api-contract.json\n    - cat dist/api-contract.json\n  artifacts:\n    paths:\n      - dist/api-contract.json\n    expire_in: 1 hour\n\ntest-frontend:\n  stage: test\n  script:\n    - echo \"All frontend tests passed!\"\n\ntrigger-backend-pipeline:\n  stage: trigger-backend\n  trigger:\n    project: my-org/backend-service\n    branch: main\n    strategy: depend\n  rules:\n    - if: $CI_COMMIT_BRANCH == \"main\"\n```\n\n```yaml\n# backend repo: .gitlab-ci.yml\nstages:\n  - build\n  - test\n\nbuild-backend:\n  stage: build\n  script:\n    - echo \"All backend tests passed!\"\n\nintegration-test:\n  stage: test\n  rules:\n    - if: $CI_PIPELINE_SOURCE == \"pipeline\"\n  script:\n    - echo \"Fetching API contract from frontend...\"\n    - |\n      curl --silent --fail \\\n        --header \"JOB-TOKEN: $CI_JOB_TOKEN\" \\\n        --output api-contract.json \\\n        \"${CI_API_V4_URL}/projects/${FRONTEND_PROJECT_ID}/jobs/artifacts/main/raw/dist/api-contract.json?job=build-frontend\"\n    - cat api-contract.json\n    - |\n      if grep -q '\"breaking_changes\": true' api-contract.json; then\n        echo \"FAIL: Breaking API changes detected - backend integration blocked!\"\n        exit 1\n      fi\n      echo \"PASS: API contract is compatible!\"\n```\n\n\nA few things worth noting in this config. The `integration-test` job uses `$CI_PIPELINE_SOURCE == \"pipeline\"` to ensure it only runs when triggered by an upstream pipeline, not on a standalone push to the backend repo. The frontend project ID is referenced via `$FRONTEND_PROJECT_ID`, which should be set as a [CI/CD variable](https://docs.gitlab.com/ci/variables/) in the backend project settings to avoid hardcoding it.\n\n\nWhy it matters: Cross-service breakage that previously surfaced in production gets caught in the pipeline instead. The dependency between services stops being invisible and becomes something teams can see, track, and act on.\n\n\n![Cross-project pipelines](https://res.cloudinary.com/about-gitlab-com/image/upload/v1775738762/Blog/Imported/hackathon-fake-blog-post-s/image4_h6mfsb.png \"Cross-project pipelines\")\n\n\n## 3. Multi-tenant / matrix deployments: Dynamic child pipelines\n\n\nThe problem: You deploy the same application to 15 customer environments, or three cloud regions, or dev/staging/prod. Updating a deploy stage across all of them one by one is the kind of work that leads to configuration drift. Writing a separate pipeline for each environment is unmaintainable from day one.\n\n\nGitLab's [dynamic child pipelines](https://docs.gitlab.com/ci/pipelines/downstream_pipelines/#dynamic-child-pipelines) let you generate a pipeline at runtime. A job runs a script that produces a YAML file, and that YAML becomes the pipeline for the next stage. The pipeline structure itself becomes data.\n\n\n```yaml\n# .gitlab-ci.yml\nstages:\n  - generate\n  - trigger-environments\n\ngenerate-config:\n  stage: generate\n  script:\n    - |\n      # ENVIRONMENTS can be passed as a CI variable or read from a config file.\n      # Default to dev, staging, prod if not set.\n      ENVIRONMENTS=${ENVIRONMENTS:-\"dev staging prod\"}\n      for ENV in $ENVIRONMENTS; do\n        cat > ${ENV}-pipeline.yml \u003C\u003C EOF\n      stages:\n        - deploy\n        - verify\n      deploy-${ENV}:\n        stage: deploy\n        script:\n          - echo \"Deploying to ${ENV} environment\"\n      verify-${ENV}:\n        stage: verify\n        script:\n          - echo \"Running smoke tests on ${ENV}\"\n      EOF\n      done\n  artifacts:\n    paths:\n      - \"*.yml\"\n    exclude:\n      - \".gitlab-ci.yml\"\n\n.trigger-template:\n  stage: trigger-environments\n  trigger:\n    strategy: depend\n\ntrigger-dev:\n  extends: .trigger-template\n  trigger:\n    include:\n      - artifact: dev-pipeline.yml\n        job: generate-config\n\ntrigger-staging:\n  extends: .trigger-template\n  needs: [trigger-dev]\n  trigger:\n    include:\n      - artifact: staging-pipeline.yml\n        job: generate-config\n\ntrigger-prod:\n  extends: .trigger-template\n  needs: [trigger-staging]\n  trigger:\n    include:\n      - artifact: prod-pipeline.yml\n        job: generate-config\n  when: manual\n```\n\n\nThe generation script loops over an `ENVIRONMENTS` variable rather than hardcoding each environment separately. Pass in a different list via a CI variable or read it from a config file and the pipeline adapts without touching the YAML. The trigger jobs use [extends:](https://docs.gitlab.com/ci/yaml/#extends) to inherit shared configuration from `.trigger-template`, so `strategy: depend` is defined once rather than repeated on every trigger job. Add a new environment by updating the variable, not by duplicating pipeline config. Add [when: manual](https://docs.gitlab.com/ci/yaml/#when) to the production trigger and you get a promotion gate baked right into the pipeline graph.\n\n\nWhy it matters: SaaS companies and platform teams use this pattern to manage dozens of environments without duplicating pipeline logic. The pipeline structure itself stays lean as the deployment matrix grows.\n\n\n![Dynamic pipeline](https://res.cloudinary.com/about-gitlab-com/image/upload/v1775738765/Blog/Imported/hackathon-fake-blog-post-s/image7_wr0kx2.png \"Dynamic pipeline\")\n\n\n## 4. MR-first delivery: Merge request pipelines, merged results, and workflow routing\n\n\nThe problem: Your pipeline runs on every push to every branch. Expensive tests run on feature branches that will never merge. Meanwhile, you have no guarantee that what you tested is actually what will land on `main` after a merge.\n\n\nGitLab has three interlocking features that solve this together:\n\n\n*   [Merge request pipelines](https://docs.gitlab.com/ci/pipelines/merge_request_pipelines/) run only when a merge request exists, not on every branch push. This alone eliminates a significant amount of wasted compute.\n\n*   [Merged results pipelines](https://docs.gitlab.com/ci/pipelines/merged_results_pipelines/) go further. GitLab creates a temporary merge commit (your branch plus the current target branch) and runs the pipeline against that. You are testing what will actually exist after the merge, not just your branch in isolation.\n\n*   [Workflow rules](https://docs.gitlab.com/ci/yaml/workflow/) let you define exactly which pipeline type runs under which conditions and suppress everything else. The `$CI_OPEN_MERGE_REQUESTS` guard below prevents duplicate pipelines firing for both a branch and its open MR simultaneously.\n\n\nWith those three working together, here is what a tiered pipeline looks like:\n\n```yaml\n# .gitlab-ci.yml\nworkflow:\n  rules:\n    - if: $CI_PIPELINE_SOURCE == \"merge_request_event\"\n    - if: $CI_COMMIT_BRANCH && $CI_OPEN_MERGE_REQUESTS\n      when: never\n    - if: $CI_COMMIT_BRANCH\n    - if: $CI_PIPELINE_SOURCE == \"schedule\"\n\nstages:\n  - fast-checks\n  - expensive-tests\n  - deploy\n\nlint-code:\n  stage: fast-checks\n  script:\n    - echo \"Running linter\"\n  rules:\n    - if: $CI_PIPELINE_SOURCE == \"push\"\n    - if: $CI_PIPELINE_SOURCE == \"merge_request_event\"\n    - if: $CI_COMMIT_BRANCH == \"main\"\n\nunit-tests:\n  stage: fast-checks\n  script:\n    - echo \"Running unit tests\"\n  rules:\n    - if: $CI_PIPELINE_SOURCE == \"push\"\n    - if: $CI_PIPELINE_SOURCE == \"merge_request_event\"\n    - if: $CI_COMMIT_BRANCH == \"main\"\n\nintegration-tests:\n  stage: expensive-tests\n  script:\n    - echo \"Running integration tests (15 min)\"\n  rules:\n    - if: $CI_PIPELINE_SOURCE == \"merge_request_event\"\n    - if: $CI_COMMIT_BRANCH == \"main\"\n\ne2e-tests:\n  stage: expensive-tests\n  script:\n    - echo \"Running E2E tests (30 min)\"\n  rules:\n    - if: $CI_PIPELINE_SOURCE == \"merge_request_event\"\n    - if: $CI_COMMIT_BRANCH == \"main\"\n\nnightly-comprehensive-scan:\n  stage: expensive-tests\n  script:\n    - echo \"Running full nightly suite (2 hours)\"\n  rules:\n    - if: $CI_PIPELINE_SOURCE == \"schedule\"\n\ndeploy-production:\n  stage: deploy\n  script:\n    - echo \"Deploying to production\"\n  rules:\n    - if: $CI_COMMIT_BRANCH == \"main\"\n      when: manual\n```\n\nWith this setup, the pipeline behaves differently depending on context. A push to a feature branch with no open MR runs lint and unit tests only. Once an MR is opened, the workflow rules switch from a branch pipeline to an MR pipeline, and the full integration and E2E suite runs against the merged result. Merging to `main` queues a manual production deployment. A nightly schedule runs the comprehensive scan once, not on every commit.\n\n\nWhy it matters: Teams routinely cut CI costs significantly with this pattern, not by running fewer tests, but by running the right tests at the right time. Merged results pipelines catch the class of bugs that only appear after a merge, before they ever reach `main`.\n\n\n![Conditional pipelines (within a branch with no MR)](https://res.cloudinary.com/about-gitlab-com/image/upload/v1775738768/Blog/Imported/hackathon-fake-blog-post-s/image6_dnfcny.png \"Conditional pipelines (within a branch with no MR)\")\n\n\n\n![Conditional pipelines (within an MR)](https://res.cloudinary.com/about-gitlab-com/image/upload/v1775738772/Blog/Imported/hackathon-fake-blog-post-s/image1_wyiafu.png \"Conditional pipelines (within an MR)\")\n\n\n\n![Conditional pipelines (on the main branch)](https://res.cloudinary.com/about-gitlab-com/image/upload/v1775738774/Blog/Imported/hackathon-fake-blog-post-s/image5_r6lkfd.png \"Conditional pipelines (on the main branch)\")\n\n## 5. Governed pipelines: CI/CD Components\n\n\nThe problem: Your platform team has defined the right way to build, test, and deploy. But every team has their own `.gitlab-ci.yml` with subtle variations. Security scanning gets skipped. Deployment standards drift. Audits are painful.\n\n\nGitLab [CI/CD Components](https://docs.gitlab.com/ci/components/) let platform teams publish versioned, reusable pipeline building blocks. Application teams consume them with a single `include:` line and optional inputs — no copy-paste, no drift. Components are discoverable through the [CI/CD Catalog](https://docs.gitlab.com/ci/components/#cicd-catalog), which means teams can find and adopt approved building blocks without needing to go through the platform team directly.\n\n\nHere is a component definition from a shared library:\n\n```yaml\n# templates/deploy.yml\nspec:\n  inputs:\n    stage:\n      default: deploy\n    environment:\n      default: production\n---\ndeploy-job:\n  stage: $[[ inputs.stage ]]\n  script:\n    - echo \"Deploying $APP_NAME to $[[ inputs.environment ]]\"\n    - echo \"Deploy URL: $DEPLOY_URL\"\n  environment:\n    name: $[[ inputs.environment ]]\n```\nAnd here is how an application team consumes it:\n\n```yaml\n# Application repo: .gitlab-ci.yml\nvariables:\n  APP_NAME: \"my-awesome-app\"\n  DEPLOY_URL: \"https://api.example.com\"\n\ninclude:\n  - component: gitlab.com/my-org/component-library/build@v1.0.6\n  - component: gitlab.com/my-org/component-library/test@v1.0.6\n  - component: gitlab.com/my-org/component-library/deploy@v1.0.6\n    inputs:\n      environment: staging\n\nstages:\n  - build\n  - test\n  - deploy\n```\n\nThree lines of `include:` replace hundreds of lines of duplicated YAML. The platform team can push a security fix to `v1.0.7` and teams opt in on their own schedule — or the platform team can pin everyone to a minimum version. Either way, one change propagates everywhere instead of needing to be applied repo by repo.\n\n\nPair this with [resource groups](https://docs.gitlab.com/ci/resource_groups/) to prevent concurrent deployments to the same environment, and [protected environments](https://docs.gitlab.com/ci/environments/protected_environments/) to enforce approval gates - and you have a governed delivery platform where compliance is the default, not the exception.\n\n\nWhy it matters: This is the pattern that makes GitLab CI/CD scale across hundreds of teams. Platform engineering teams enforce compliance without becoming a bottleneck. Application teams get a fast path to a working pipeline without reinventing the wheel.\n\n\n![Component pipeline (imported jobs)](https://res.cloudinary.com/about-gitlab-com/image/upload/v1775738776/Blog/Imported/hackathon-fake-blog-post-s/image2_pizuxd.png \"Component pipeline (imported jobs)\")\n\n## Putting it all together\n\nNone of these features exist in isolation. The reason GitLab's pipeline model is worth understanding deeply is that these primitives compose:\n\n*   A monorepo uses parent-child pipelines, and each child uses DAG execution\n\n*   A microservices platform uses multi-project pipelines, and each project uses MR pipelines with merged results\n\n*   A governed platform uses CI/CD components to standardize the patterns above across every team\n\n\nMost teams discover one of these features when they hit a specific pain point. The ones who invest in understanding the full model end up with a delivery system that actually reflects how their engineering organization works, not a pipeline that fights it.\n\n## Other patterns worth exploring\n\n\nThe five patterns above cover the most common structural pain points, but GitLab's pipeline model goes further. A few others worth looking into as your needs grow:\n\n\n*   [Review apps with dynamic environments](https://docs.gitlab.com/ci/environments/) let you spin up a live preview for every feature branch and tear it down automatically when the MR closes. Useful for teams doing frontend work or API changes that need stakeholder sign-off before merging.\n\n*   [Caching and artifact strategies](https://docs.gitlab.com/ci/caching/) are often the fastest way to cut pipeline runtime after the structural work is done. Structuring `cache:` keys around dependency lockfiles and being deliberate about what gets passed between jobs with [artifacts:](https://docs.gitlab.com/ci/yaml/#artifacts) can make a significant difference without changing your pipeline shape at all.\n\n*   [Scheduled and API-triggered pipelines](https://docs.gitlab.com/ci/pipelines/schedules/) are worth knowing about because not everything should run on a code push. Nightly security scans, compliance reports, and release automation are better modeled as scheduled or [API-triggered](https://docs.gitlab.com/ci/triggers/) pipelines with `$CI_PIPELINE_SOURCE` routing the right jobs for each context.\n\n## How to get started\n\nModern software delivery is complex. Teams are managing monorepos with dozens of services, coordinating across multiple repositories, deploying to many environments at once, and trying to keep standards consistent as organizations grow. GitLab's pipeline model was built with all of that in mind.\n\nWhat makes it worth investing time in is how well the pieces fit together. Parent-child pipelines bring structure to large codebases. Multi-project pipelines make cross-team dependencies visible and testable. Dynamic pipelines turn environment management into something that scales gracefully. MR-first delivery with merged results ensures confidence at every step of the review process. And CI/CD Components give platform teams a way to share best practices across an entire organization without becoming a bottleneck.\n\nEach of these features is powerful on its own, and even more so when combined. GitLab gives you the building blocks to design a delivery system that fits how your team actually works, and grows with you as your needs evolve.\n\n> [Start a free trial of GitLab Ultimate](https://about.gitlab.com/free-trial/) to use pipeline logic today.\n\n## Read more\n\n*   [Variable and artifact sharing in GitLab parent-child pipelines](https://about.gitlab.com/blog/variable-and-artifact-sharing-in-gitlab-parent-child-pipelines/)\n*   [CI/CD inputs: Secure and preferred method to pass parameters to a pipeline](https://about.gitlab.com/blog/ci-cd-inputs-secure-and-preferred-method-to-pass-parameters-to-a-pipeline/)\n*   [Tutorial: How to set up your first GitLab CI/CD component](https://about.gitlab.com/blog/tutorial-how-to-set-up-your-first-gitlab-ci-cd-component/)\n*   [How to include file references in your CI/CD components](https://about.gitlab.com/blog/how-to-include-file-references-in-your-ci-cd-components/)\n*   [FAQ: GitLab CI/CD Catalog](https://about.gitlab.com/blog/faq-gitlab-ci-cd-catalog/)\n*   [Building a GitLab CI/CD pipeline for a monorepo the easy way](https://about.gitlab.com/blog/building-a-gitlab-ci-cd-pipeline-for-a-monorepo-the-easy-way/)\n*   [A CI/CD component builder's journey](https://about.gitlab.com/blog/a-ci-component-builders-journey/)\n*   [CI/CD Catalog goes GA: No more building pipelines from scratch](https://about.gitlab.com/blog/ci-cd-catalog-goes-ga-no-more-building-pipelines-from-scratch/)","5 ways GitLab pipeline logic solves real engineering problems","Learn how to scale CI/CD with composable patterns for monorepos, microservices, environments, and governance.",[720],"Omid Khan","https://res.cloudinary.com/about-gitlab-com/image/upload/v1772721753/frfsm1qfscwrmsyzj1qn.png","2026-04-09",[102,724,725,726],"DevOps platform","tutorial","features",{"featured":24,"template":13,"slug":728},"5-ways-gitlab-pipeline-logic-solves-real-engineering-problems",{"content":730,"config":740},{"title":731,"description":732,"authors":733,"heroImage":735,"date":736,"body":737,"category":9,"tags":738},"How to use GitLab Container Virtual Registry with Docker Hardened Images","Learn how to simplify container image management with this step-by-step guide.",[734],"Tim Rizzi","https://res.cloudinary.com/about-gitlab-com/image/upload/v1772111172/mwhgbjawn62kymfwrhle.png","2026-03-12","If you're a platform engineer, you've probably had this conversation:\n  \n*\"Security says we need to use hardened base images.\"*\n\n*\"Great, where do I configure credentials for yet another registry?\"*\n\n*\"Also, how do we make sure everyone actually uses them?\"*\n\nOr this one:\n\n*\"Why are our builds so slow?\"*\n\n*\"We're pulling the same 500MB image from Docker Hub in every single job.\"*\n\n*\"Can't we just cache these somewhere?\"*\n\nI've been working on [Container Virtual Registry](https://docs.gitlab.com/user/packages/virtual_registry/container/) at GitLab specifically to solve these problems. It's a pull-through cache that sits in front of your upstream registries — Docker Hub, dhi.io (Docker Hardened Images), MCR, and Quay — and gives your teams a single endpoint to pull from. Images get cached on the first pull. Subsequent pulls come from the cache. Your developers don't need to know or care which upstream a particular image came from.\n\nThis article shows you how to set up Container Virtual Registry, specifically with Docker Hardened Images in mind, since that's a combination that makes a lot of sense for teams concerned about security and not making their developers' lives harder.\n\n## What problem are we actually solving?\n\nThe Platform teams I usually talk to manage container images across three to five registries:\n\n* **Docker Hub** for most base images\n* **dhi.io** for Docker Hardened Images (security-conscious workloads)\n* **MCR** for .NET and Azure tooling\n* **Quay.io** for Red Hat ecosystem stuff\n* **Internal registries** for proprietary images\n\nEach one has its own:\n\n* Authentication mechanism\n* Network latency characteristics\n* Way of organizing image paths\n\nYour CI/CD configs end up littered with registry-specific logic. Credential management becomes a project unto itself. And every pipeline job pulls the same base images over the network, even though they haven't changed in weeks.\n\nContainer Virtual Registry consolidates this. One registry URL. One authentication flow (GitLab's). Cached images are served from GitLab's infrastructure rather than traversing the internet each time.\n\n## How it works\n\nThe model is straightforward:\n\n```text\nYour pipeline pulls:\n  gitlab.com/virtual_registries/container/1000016/python:3.13\n\nVirtual registry checks:\n  1. Do I have this cached? → Return it\n  2. No? → Fetch from upstream, cache it, return it\n\n```\n\nYou configure upstreams in priority order. When a pull request comes in, the virtual registry checks each upstream until it finds the image. The result gets cached for a configurable period (default 24 hours).\n\n```text\n┌─────────────────────────────────────────────────────────┐\n│                    CI/CD Pipeline                       │\n│                          │                              │\n│                          ▼                              │\n│   gitlab.com/virtual_registries/container/\u003Cid>/image   │\n└─────────────────────────────────────────────────────────┘\n                           │\n                           ▼\n┌─────────────────────────────────────────────────────────┐\n│            Container Virtual Registry                   │\n│                                                         │\n│  Upstream 1: Docker Hub ────────────────┐               │\n│  Upstream 2: dhi.io (Hardened) ────────┐│               │\n│  Upstream 3: MCR ─────────────────────┐││               │\n│  Upstream 4: Quay.io ────────────────┐│││               │\n│                                      ││││               │\n│                    ┌─────────────────┴┴┴┴──┐            │\n│                    │        Cache          │            │\n│                    │  (manifests + layers) │            │\n│                    └───────────────────────┘            │\n└─────────────────────────────────────────────────────────┘\n```\n\n## Why this matters for Docker Hardened Images\n\n[Docker Hardened Images](https://docs.docker.com/dhi/) are great because of the minimal attack surface, near-zero CVEs, proper software bills of materials (SBOMs), and SLSA provenance. If you're evaluating base images for security-sensitive workloads, they should be on your list.\n\nBut adopting them creates the same operational friction as any new registry:\n\n* **Credential distribution**: You need to get Docker credentials to every system that pulls images from dhi.io.\n* **CI/CD changes**: Every pipeline needs to be updated to authenticate with dhi.io.\n* **Developer friction**: People need to remember to use the hardened variants.\n* **Visibility gap**: It's difficult to tell if teams are actually using hardened images vs. regular ones.\n\nVirtual registry addresses each of these:\n\n**Single credential**: Teams authenticate to GitLab. The virtual registry handles upstream authentication. You configure Docker credentials once, at the registry level, and they apply to all pulls.\n\n**No CI/CD changes per-team**: Point pipelines at your virtual registry. Done. The upstream configuration is centralized.\n\n**Gradual adoption**: Since images get cached with their full path, you can see in the cache what's being pulled. If someone's pulling `library/python:3.11` instead of the hardened variant, you'll know.\n\n**Audit trail**: The cache shows you exactly which images are in active use. Useful for compliance, useful for understanding what your fleet actually depends on.\n\n## Setting it up\n\nHere's a real setup using the Python client from this demo project.\n\n### Create the virtual registry\n\n```python\nfrom virtual_registry_client import VirtualRegistryClient\n\nclient = VirtualRegistryClient()\n\nregistry = client.create_virtual_registry(\n    group_id=\"785414\",  # Your top-level group ID\n    name=\"platform-images\",\n    description=\"Cached container images for platform teams\"\n)\n\nprint(f\"Registry ID: {registry['id']}\")\n# You'll need this ID for the pull URL\n```\n\n### Add Docker Hub as an upstream\n\nFor official images like Alpine, Python, etc.:\n\n```python\ndocker_upstream = client.create_upstream(\n    registry_id=registry['id'],\n    url=\"https://registry-1.docker.io\",\n    name=\"Docker Hub\",\n    cache_validity_hours=24\n)\n```\n\n### Add Docker Hardened Images (dhi.io)\n\nDocker Hardened Images are hosted on `dhi.io`, a separate registry that requires authentication:\n\n```python\ndhi_upstream = client.create_upstream(\n    registry_id=registry['id'],\n    url=\"https://dhi.io\",\n    name=\"Docker Hardened Images\",\n    username=\"your-docker-username\",\n    password=\"your-docker-access-token\",\n    cache_validity_hours=24\n)\n```\n\n### Add other upstreams\n\n```python\n# MCR for .NET teams\nclient.create_upstream(\n    registry_id=registry['id'],\n    url=\"https://mcr.microsoft.com\",\n    name=\"Microsoft Container Registry\",\n    cache_validity_hours=48\n)\n\n# Quay for Red Hat stuff\nclient.create_upstream(\n    registry_id=registry['id'],\n    url=\"https://quay.io\",\n    name=\"Quay.io\",\n    cache_validity_hours=24\n)\n```\n\n### Update your CI/CD\n\nHere's a `.gitlab-ci.yml` that pulls through the virtual registry:\n\n```yaml\nvariables:\n  VIRTUAL_REGISTRY_ID: \u003Cyour_virtual_registry_ID>\n\n  \nbuild:\n  image: docker:24\n  services:\n    - docker:24-dind\n  before_script:\n    # Authenticate to GitLab (which handles upstream auth for you)\n    - echo \"${CI_JOB_TOKEN}\" | docker login -u gitlab-ci-token --password-stdin gitlab.com\n  script:\n    # All of these go through your single virtual registry\n    \n    # Official Docker Hub images (use library/ prefix)\n    - docker pull gitlab.com/virtual_registries/container/${VIRTUAL_REGISTRY_ID}/library/alpine:latest\n    \n    # Docker Hardened Images from dhi.io (no prefix needed)\n    - docker pull gitlab.com/virtual_registries/container/${VIRTUAL_REGISTRY_ID}/python:3.13\n    \n    # .NET from MCR\n    - docker pull gitlab.com/virtual_registries/container/${VIRTUAL_REGISTRY_ID}/dotnet/sdk:8.0\n```\n\n### Image path formats\n\nDifferent registries use different path conventions:\n\n| Registry | Pull URL Example |\n|----------|------------------|\n| Docker Hub (official) | `.../library/python:3.11-slim` |\n| Docker Hardened Images (dhi.io) | `.../python:3.13` |\n| MCR | `.../dotnet/sdk:8.0` |\n| Quay.io | `.../prometheus/prometheus:latest` |\n\n### Verify it's working\n\nAfter some pulls, check your cache:\n\n```python\nupstreams = client.list_registry_upstreams(registry['id'])\nfor upstream in upstreams:\n    entries = client.list_cache_entries(upstream['id'])\n    print(f\"{upstream['name']}: {len(entries)} cached entries\")\n\n```\n\n## What the numbers look like\n\nI ran tests pulling images through the virtual registry:\n\n| Metric | Without Cache | With Warm Cache |\n|--------|---------------|-----------------|\n| Pull time (Alpine) | 10.3s | 4.2s |\n| Pull time (Python 3.13 DHI) | 11.6s | ~4s |\n| Network roundtrips to upstream | Every pull | Cache misses only |\n\n\n\n\nThe first pull is the same speed (it has to fetch from upstream). Every pull after that, for the cache validity period, comes straight from GitLab's storage. No network hop to Docker Hub, dhi.io, MCR, or wherever the image lives.\n\nFor a team running hundreds of pipeline jobs per day, that's hours of cumulative build time saved.\n\n## Practical considerations\nHere are some considerations to keep in mind:\n\n### Cache validity\n\n24 hours is the default. For security-sensitive images where you want patches quickly, consider 12 hours or less:\n\n```python\nclient.create_upstream(\n    registry_id=registry['id'],\n    url=\"https://dhi.io\",\n    name=\"Docker Hardened Images\",\n    username=\"your-username\",\n    password=\"your-token\",\n    cache_validity_hours=12\n)\n```\n\nFor stable, infrequently-updated images (like specific version tags), longer validity is fine.\n\n### Upstream priority\n\nUpstreams are checked in order. If you have images with the same name on different registries, the first matching upstream wins.\n\n### Limits\n\n* Maximum of 20 virtual registries per group\n* Maximum of 20 upstreams per virtual registry\n\n## Configuration via UI\n\nYou can also configure virtual registries and upstreams directly from the GitLab UI—no API calls required. Navigate to your group's **Settings > Packages and registries > Virtual Registry** to:\n\n* Create and manage virtual registries\n* Add, edit, and reorder upstream registries\n* View and manage the cache\n* Monitor which images are being pulled\n\n## What's next\n\nWe're actively developing:\n\n* **Allow/deny lists**: Use regex to control which images can be pulled from specific upstreams.\n\nThis is beta software. It works, people are using it in production, but we're still iterating based on feedback.\n\n## Share your feedback\n\nIf you're a platform engineer dealing with container registry sprawl, I'd like to understand your setup:\n\n* How many upstream registries are you managing?\n* What's your biggest pain point with the current state?\n* Would something like this help, and if not, what's missing?\n\nPlease share your experiences in the [Container Virtual Registry feedback issue](https://gitlab.com/gitlab-org/gitlab/-/work_items/589630).\n## Related resources\n- [New GitLab metrics and registry features help reduce CI/CD bottlenecks](https://about.gitlab.com/blog/new-gitlab-metrics-and-registry-features-help-reduce-ci-cd-bottlenecks/#container-virtual-registry)\n- [Container Virtual Registry documentation](https://docs.gitlab.com/user/packages/virtual_registry/container/)\n- [Container Virtual Registry API](https://docs.gitlab.com/api/container_virtual_registries/)",[725,739,726],"product",{"featured":12,"template":13,"slug":741},"using-gitlab-container-virtual-registry-with-docker-hardened-images",{"content":743,"config":753},{"title":744,"description":745,"authors":746,"heroImage":748,"date":749,"category":9,"tags":750,"body":752},"How IIT Bombay students are coding the future with GitLab","At GitLab, we often talk about how software accelerates innovation. But sometimes, you have to step away from the Zoom calls and stand in a crowded university hall to remember why we do this.",[747],"Nick Veenhof","https://res.cloudinary.com/about-gitlab-com/image/upload/v1750099013/Blog/Hero%20Images/Blog/Hero%20Images/blog-image-template-1800x945%20%2814%29_6VTUA8mUhOZNDaRVNPeKwl_1750099012960.png","2026-01-08",[255,617,751],"open source","The GitLab team recently had the privilege of judging the **iHack Hackathon** at **IIT Bombay's E-Summit**. The energy was electric, the coffee was flowing, and the talent was undeniable. But what struck us most wasn't just the code — it was the sheer determination of students to solve real-world problems, often overcoming significant logistical and financial hurdles to simply be in the room.\n\n\nThrough our [GitLab for Education program](https://about.gitlab.com/solutions/education/), we aim to empower the next generation of developers with tools and opportunity. Here is a look at what the students built, and how they used GitLab to bridge the gap between idea and reality.\n\n## The challenge: Build faster, build securely\n\nThe premise for the GitLab track of the hackathon was simple: Don't just show us a product; show us how you built it. We wanted to see how students utilized GitLab's platform — from Issue Boards to CI/CD pipelines — to accelerate the development lifecycle.\n\nThe results were inspiring.\n\n## The winners\n\n### 1st place: Team Decode — Democratizing Scientific Research\n\n**Project:** FIRE (Fast Integrated Research Environment)\n\nTeam Decode took home the top prize with a solution that warms a developer's heart: a local-first, blazing-fast data processing tool built with [Rust](https://about.gitlab.com/blog/secure-rust-development-with-gitlab/) and Tauri. They identified a massive pain point for data science students: existing tools are fragmented, slow, and expensive.\n\nTheir solution, FIRE, allows researchers to visualize complex formats (like NetCDF) instantly. What impressed the judges most was their \"hacker\" ethos. They didn't just build a tool; they built it to be open and accessible.\n\n**How they used GitLab:** Since the team lived far apart, asynchronous communication was key. They utilized **GitLab Issue Boards** and **Milestones** to track progress and integrated their repo with Telegram to get real-time push notifications. As one team member noted, \"Coordinating all these technologies was really difficult, and what helped us was GitLab... the Issue Board really helped us track who was doing what.\"\n\n![Team Decode](https://res.cloudinary.com/about-gitlab-com/image/upload/v1767380253/epqazj1jc5c7zkgqun9h.jpg)\n\n### 2nd place: Team BichdeHueDost — Reuniting to Solve Payments\n\n**Project:** SemiPay (RFID Cashless Payment for Schools)\n\nThe team name, BichdeHueDost, translates to \"Friends who have been set apart.\" It's a fitting name for a group of friends who went to different colleges but reunited to build this project. They tackled a unique problem: handling cash in schools for young children. Their solution used RFID cards backed by a blockchain ledger to ensure secure, cashless transactions for students.\n\n**How they used GitLab:** They utilized [GitLab CI/CD](https://about.gitlab.com/topics/ci-cd/) to automate the build process for their Flutter application (APK), ensuring that every commit resulted in a testable artifact. This allowed them to iterate quickly despite the \"flaky\" nature of cross-platform mobile development.\n\n![Team BichdeHueDost](https://res.cloudinary.com/about-gitlab-com/image/upload/v1767380253/pkukrjgx2miukb6nrj5g.jpg)\n\n### 3rd place: Team ZenYukti — Agentic Repository Intelligence\n\n**Project:** RepoInsight AI (AI-powered, GitLab-native intelligence platform)\n\nTeam ZenYukti impressed us with a solution that tackles a universal developer pain point: understanding unfamiliar codebases. What stood out to the judges was the tool's practical approach to onboarding and code comprehension: RepoInsight-AI automatically generates documentation, visualizes repository structure, and even helps identify bugs, all while maintaining context about the entire codebase.\n\n**How they used GitLab:** The team built a comprehensive CI/CD pipeline that showcased GitLab's security and DevOps capabilities. They integrated [GitLab's Security Templates](https://gitlab.com/gitlab-org/gitlab/-/tree/master/lib/gitlab/ci/templates/Security) (SAST, Dependency Scanning, and Secret Detection), and utilized [GitLab Container Registry](https://docs.gitlab.com/user/packages/container_registry/) to manage their Docker images for backend and frontend components. They created an AI auto-review bot that runs on merge requests, demonstrating an \"agentic workflow\" where AI assists in the development process itself.\n\n![Team ZenYukti](https://res.cloudinary.com/about-gitlab-com/image/upload/v1767380253/ymlzqoruv5al1secatba.jpg)\n\n## Beyond the code: A lesson in inclusion\n\nWhile the code was impressive, the most powerful moment of the event happened away from the keyboard.\n\nDuring the feedback session, we learned about the journey Team ZenYukti took to get to Mumbai. They traveled over 24 hours, covering nearly 1,800 kilometers. Because flights were too expensive and trains were booked, they traveled in the \"General Coach,\" a non-reserved, severely overcrowded carriage.\n\nAs one student described it:\n\n*\"You cannot even imagine something like this... there are no seats... people sit on the top of the train. This is what we have endured.\"*\n\nThis hit home. [Diversity, Inclusion, and Belonging](https://handbook.gitlab.com/handbook/company/culture/inclusion/) are core values at GitLab. We realized that for these students, the barrier to entry wasn't intellect or skill, it was access.\n\nIn that moment, we decided to break that barrier. We committed to reimbursing the travel expenses for the participants who struggled to get there. It's a small step, but it underlines a massive truth: **talent is distributed equally, but opportunity is not.**\n\n![hackathon class together](https://res.cloudinary.com/about-gitlab-com/image/upload/v1767380252/o5aqmboquz8ehusxvgom.jpg)\n\n### The future is bright (and automated)\n\nWe also saw incredible potential in teams like Prometheus, who attempted to build an autonomous patch remediation tool (DevGuardian), and Team Arrakis, who built a voice-first job portal for blue-collar workers using [GitLab Duo](https://about.gitlab.com/gitlab-duo-agent-platform/) to troubleshoot their pipelines.\n\nTo all the students who participated: You are the future. Through [GitLab for Education](https://about.gitlab.com/solutions/education/), we are committed to providing you with the top-tier tools (like GitLab Ultimate) you need to learn, collaborate, and change the world — whether you are coding from a dorm room, a lab, or a train carriage. **Keep shipping.**\n\n> :bulb: Learn more about the [GitLab for Education program](https://about.gitlab.com/solutions/education/).\n",{"slug":754,"featured":12,"template":13},"how-iit-bombay-students-code-future-with-gitlab",{"promotions":756},[757,771,782,794],{"id":758,"categories":759,"header":761,"text":762,"button":763,"image":768},"ai-modernization",[760],"ai-ml","Is AI achieving its promise at scale?","Quiz will take 5 minutes or less",{"text":764,"config":765},"Get your AI maturity score",{"href":766,"dataGaName":767,"dataGaLocation":237},"/assessments/ai-modernization-assessment/","modernization assessment",{"config":769},{"src":770},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1772138786/qix0m7kwnd8x2fh1zq49.png",{"id":772,"categories":773,"header":774,"text":762,"button":775,"image":779},"devops-modernization",[739,563],"Are you just managing tools or shipping innovation?",{"text":776,"config":777},"Get your DevOps maturity score",{"href":778,"dataGaName":767,"dataGaLocation":237},"/assessments/devops-modernization-assessment/",{"config":780},{"src":781},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1772138785/eg818fmakweyuznttgid.png",{"id":783,"categories":784,"header":786,"text":762,"button":787,"image":791},"security-modernization",[785],"security","Are you trading speed for security?",{"text":788,"config":789},"Get your security maturity score",{"href":790,"dataGaName":767,"dataGaLocation":237},"/assessments/security-modernization-assessment/",{"config":792},{"src":793},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1772138786/p4pbqd9nnjejg5ds6mdk.png",{"id":795,"paths":796,"header":799,"text":800,"button":801,"image":806},"github-azure-migration",[797,798],"migration-from-azure-devops-to-gitlab","integrating-azure-devops-scm-and-gitlab","Is your team ready for GitHub's Azure move?","GitHub is already rebuilding around Azure. Find out what it means for you.",{"text":802,"config":803},"See how GitLab compares to GitHub",{"href":804,"dataGaName":805,"dataGaLocation":237},"/compare/gitlab-vs-github/github-azure-migration/","github azure migration",{"config":807},{"src":781},{"header":809,"blurb":810,"button":811,"secondaryButton":816},"Start building faster today","See what your team can do with the intelligent orchestration platform for DevSecOps.\n",{"text":812,"config":813},"Get your free trial",{"href":814,"dataGaName":44,"dataGaLocation":815},"https://gitlab.com/-/trial_registrations/new?glm_content=default-saas-trial&glm_source=about.gitlab.com/","feature",{"text":499,"config":817},{"href":48,"dataGaName":49,"dataGaLocation":815},1776458698477]