Setting</th>	Value</th> </tr> </thead>
default_pool_size</td>	100</td> </tr>
reserve_pool_size</td>	5</td> </tr>
reserve_pool_timeout</td>	3</td> </tr>
max_client_conn</td>	2048</td> </tr>
pool_mode</td>	transaction</td> </tr>
server_idle_timeout</td>	30</td> </tr> </tbody> </table> With that all said there is still some work left to be done such as: implementing service discovery (#2042</a>), improving how we check if a secondary is available (#2866</a>), and ignoring secondaries that are too far behind the primary (#2197</a>).</p> It's worth mentioning that we currently do not have any plans of turning our load balancing solution into a standalone library that you can use outside of GitLab, instead our focus is on providing a solid load balancing solution for GitLab EE.</p> If this has gotten you interested and you enjoy working with databases, improving application performance, and adding database-related features to GitLab (such as service discovery</a>) you should definitely check out the job opening</a> and the database specialist handbook entry</a> for more information.</p> GitLab Patch Release: 9.5.6 2017-09-28T00:00:00+00:00 Today we are releasing version 9.5.6 for GitLab Community Edition (CE) and Enterprise Edition (EE).</p> This version resolves a number of regressions and bugs in this month's 9.5.5 release</a> and prior versions.</p> It includes the following fixes:</p> CE/EE</strong> Fix MR widget with ready to merge buttons/controls at mobile breakpoint (!14242</a>)</li> CE/EE</strong> Fix MR widget with external CI services/integrations (!13977</a>)</li> CE/EE</strong> Update "x/x discussions resolved" checkmark icon to be green when all discussions resolved (!14255</a>)</li> CE/EE</strong> Upgrade Nokogiri because of CVE-2017-9050 (!14427</a>)</li> CE/EE</strong> Handle error when fetching ref for MR with deleted source branch (!14550</a>)</li> EE</strong> Show group tab if member lock is enabled (!3025</a>)</li> Omnibus CE/EE</strong> Bump ruby version to 2.3.5 for security fixes (!1930</a>)</li> Omnibus CE/EE</strong> Resolve "update_databases assumes a populated databases.json" (!1924</a>)</li> Omnibus CE/EE</strong> Resolve "restart repmgrd after node register" (!1928</a>)</li> Omnibus CE/EE</strong> Resolve "repmgr unregister doesn't accept –node" (!1927</a>)</li> </ul> Upgrade barometer</h2> This version does not include any new migrations, and should not require any downtime.</p> Please be aware that by default the Omnibus packages will stop, run migrations, and start again, no matter how “big” or “small” the upgrade is. This behavior can be changed by adding a `/etc/gitlab/skip-auto-migrations</code> file</a>.</p>` `Updating</h2>To update, check out our update page</a>.</p> Enterprise Edition</h2> Interested in GitLab Enterprise Edition? Check out the features exclusive to EE</a>.</p> Access to GitLab Enterprise Edition is granted by a subscription</a>. No time to upgrade GitLab yourself? Subscribers receive upgrade and installation services.</p>` GitLab Patch Release: 10.0.2 2017-09-27T00:00:00+00:00 Today we are releasing version 10.0.2 for GitLab Community Edition (CE) and Enterprise Edition (EE).</p> This version resolves a number of regressions and bugs in this month's 10.0.1 release</a> and prior versions.</p> It includes the following fixes:</p> CE/EE:</strong> Resize comment form after note submit and discard it (!14401</a>)</li> CE/EE:</strong> Fix rendering double note issue (!14429</a>)</li> CE/EE:</strong> Ensure that Blob.raw returns always a valid blob object (!14412</a>)</li> CE/EE:</strong> Remove navbar gradient (!14457</a>)</li> CE/EE:</strong> Re-allow name</code> attribute on user-provided anchor HTML (!14452</a>)</li> CE/EE:</strong> Fix checks failing with 'undefined method run_command</code> in rake gitlab:check</code> (!14469</a>)</li> CE/EE:</strong> Fix PG::UniqueViolation external_email migration (!14474</a>)</li> CE/EE:</strong> Prevent URL concatenation for avatars (!14437</a>)</li> CE/EE:</strong> IssueNotes: Switch back to Write pane when note cancel or submit (!14402</a>)</li> CE/EE:</strong> Fix a dash being rendered in the note's access role (!14450</a>)</li> CE/EE:</strong> Make locked setting of Runner to not affect jobs scheduling (!14483</a>)</li> CE/EE:</strong> Avoid wrapping users in dropdown (!14410</a>)</li> EE:</strong> Fix delta size check to handle commit or nil objects (!2976</a>)</li> EE:</strong> Fix 500 error due to promote group Web hooks (!2972</a>)</li> EE:</strong> Add notes on authentication to PG HA document (!2851</a>)</li> EE:</strong> Send valid project path as name for Jira dev panel (!3005</a>)</li> EE:</strong> Remove duplicate group boards doc (!2991</a>)</li> EE:</strong> Point to embedded Consul (!2980</a>)</li> EE:</strong> Fix internal link in docs(!3003</a>)</li> Omnibus:</strong> Fix an issue where enabling a GitLab Geo role would also disable all default services (!1963</a>)</li> Omnibus:</strong> Add option to configure redis snapshot frequency (!1964</a>)</li> Omnibus:</strong> Update pg-upgrade output to be more clear when a bundled PostgreSQL is not in use (!1962</a>)</li> Omnibus:</strong> Reload consul on config change instead of restarting (!1966</a>)</li> </ul> Upgrade barometer</h2> This version does not include any new migrations, and should not require any downtime.</p> Please be aware that by default the Omnibus packages will stop, run migrations, and start again, no matter how “big” or “small” the upgrade is. This behavior can be changed by adding a /etc/gitlab/skip-auto-migrations</code> file</a>.</p> Updating</h2> To update, check out our update page</a>.</p> Enterprise Edition</h2> Interested in GitLab Enterprise Edition? Check out the features exclusive to EE</a>.</p> Access to GitLab Enterprise Edition is granted by a subscription</a>. No time to upgrade GitLab yourself? Subscribers receive upgrade and installation services.</p> GitLab Continuous Integration named a Leader in the Forrester Wave™ 2017-09-27T00:00:00+00:00 Today Forrester has evaluated GitLab as a Leader in Continuous Integration in The Forrester Wave™: Continuous Integration Tools, Q3 2017 report. As the report states, “GitLab delivers ease of use, scalability, integration, and innovation.” We see CI/CD as a critical component of our offering, making it quicker and easier to deliver value through automation. The progress we’ve made in developing a CI/CD solution that’s seamlessly integrated is thanks largely to the participation and feedback of our community and customers.</p> </p> We’re excited about what the future holds for GitLab CI/CD</a>, so it’s great to know that others see its value. Furthermore Forrester states, “GitLab’s vision is to serve enterprise-scale, integrated software development teams that want to spend more time writing code and less time maintaining their tool chain.”</p> We’ve said it before (we say it a lot, actually): CI/CD is the cornerstone of modern software development. Delivering what customers want, faster, requires a modernized software development lifecycle that saves time, effort, and cost. Automation is key to this more efficient release cycle, which is why we’ve spent a lot of time developing our built-in test and release automation, so developers can spend less time stringing together their tool chain, and more on innovation and integrating customer feedback. We’re gratified to have been named a leader in CI by Forrester in their report on Continuous Integration Tools for Q3, 2017, which evaluated 10 vendors across 25 different criteria.</p> </p> We had some key areas of success:</p> Top score in the Current Offering category</h2> GitLab received the highest score in Forrester’s Current Offering evaluation, which we believe confirms that our product is hard to beat when it comes to ease of installation, configuring builds, platform support, analytics, an intuitive UI, container support and more. This all adds up to a more seamless development process which facilitates DevOps practices and collaboration.</p> Highest score possible in the Strategy category</h2> Forrester assessed our product strategy, market approach, training, consulting and support, giving us a 5.0, the highest possible score, for all three criteria. This is huge validation for our CI/CD vision, shaped with the help of our customers, community, and their contributions.</p> Just the beginning</h2> A significant part of that vision, Auto DevOps, just shipped with our 10.0 release</a>. Auto DevOps is a hands-free DevOps experience that automatically configures your build, test, code quality assurance, review apps, deployment, and monitoring in a single environment. Our out-of-the-box templates allow you to set up an end-to-end DevOps lifecycle at the push of a button.</p> Auto DevOps is just the first step on a journey to building a complete and highly automated DevOps tool chain. Building on strong momentum, we’re working on a number of new features to better leverage our CI/CD tools in a more automated way for both operations as well as development teams. With the full automation we are planning, you will not only be able to deliver a new, functioning application from a developer’s machine to a production environment in minutes, but also monitor all production environments in one dashboard.</p> Faster time to value</h3> Reducing the time teams spend on manual tasks means you can focus on what matters: building great products that your users want. GitLab CI/CD facilitates a more iterative approach to software development, by automatically testing new code, helping you to catch potential problems earlier in the release cycle, taking the anxiety out of deployments and reducing the cost of fixing bugs. This puts you in a position to release more often, shortening the feedback loop and helping you to release features and improvements that your customers are asking for.</p> Collaboration without dependency</h3> The fact that GitLab CI/CD is fully integrated makes all the difference. Less context-switching means key information doesn’t slip through the cracks like it can when teams work in different tools, with no single source of truth. Working in an integrated tool fosters collaboration by improving visibility, keeping everyone on the same page and making it easy for different teams to participate and comment. Features like review apps – which automatically spin up a dynamic environment for merge requests, so you can preview changes right away – make it even easier to share improvements with stakeholders and gather feedback.</p> Being named a Leader in this Wave evaluation, and receiving the Top Score in Current Offering, as well as the highest score possible for Strategy, is validation for us of the strength of today’s GitLab CI offering as well as our vision for the future.</p> View the full report</a> to learn more about GitLab Continuous Integration.</p> GitLab Patch Release: 10.0.1 2017-09-24T00:00:00+00:00 Today we are releasing version 10.0.1 for GitLab Community Edition (CE) and Enterprise Edition (EE).</p> This version resolves a migration issue in this month's 10.0.0 release</a> that affected instances with inactive LDAP accounts</a>. Other than that, this update contains no changes to functionality from 10.0.0.</p> It includes the following fix:</p> CE/EE</strong> Fix duplicate key errors in PostDeployMigrateUserExternalMailData migration. (!14460</a>)</li> </ul> Thanks to Ian Yang for suggesting a solution.</p> Upgrade barometer</h2> This version includes a revision to a single migration, but does not require any downtime. If you have already successfully upgraded to 10.0.0, this upgrade will not affect you.</p> Please be aware that by default the Omnibus packages will stop, run migrations, and start again, no matter how “big” or “small” the upgrade is. This behavior can be changed by adding a /etc/gitlab/skip-auto-migrations</code> file</a>.</p> Updating</h2> To update, check out our update page</a>.</p> Enterprise Edition</h2> Interested in GitLab Enterprise Edition? Check out the features exclusive to EE</a>.</p> Access to GitLab Enterprise Edition is granted by a subscription</a>. No time to upgrade GitLab yourself? Subscribers receive upgrade and installation services.</p> GitLab 10.0 Released with Auto DevOps and Group Issue Boards 2017-09-22T00:00:00+00:00 From the formulation of an idea to executing and monitoring it in production, DevOps establishes a culture and environment where developing, testing, and releasing software can happen quickly, frequently, and more reliably.</p> GitLab 10.0 delivers a hands-free DevOps environment with the introduction of Auto DevOps</a>, allowing your team to easily configure and adopt modern development practices in your workflow. Not only that, there's new navigation</a> and a new way of collaborating across groups</a>.</p> With every monthly release of GitLab, we introduce new capabilities and improve our existing features. GitLab 10.0 is no exception and includes numerous new additions, such as the ability to automatically resolve outdated merge request discussions</a>, improvements to subgroups</a>, and an API for Wiki</a> thanks to a contribution from our open source community.</p> GitLab's powerful issue management capabilities keep getting better with every release. Filtering and searching issues across groups has been vastly improved</a>, our updated UX makes moving issues easier to discover</a> and can be automated through quick action commands</a>. GitLab Enterprise Edition Premium customers using JIRA can now see commits and branches</a> in JIRA's development panel.</p> Security and performance continues to improve. Administrators can now restrict SSH</a> access through technology and key length. LDAP Group Sync can be automated</a> through our API and can now lock down External Users</a> at point of login as well. Performance</a> continues to get faster, improving page loading speeds, the speed of creating projects and performing commits, and reduced memory usage.</p>

Scaling the GitLab database

2017-10-02T00:00:00+00:00

For a long time GitLab.com used a single PostgreSQL database server and a single replica for disaster recovery purposes. This worked reasonably well for the first few years of GitLab.com's existence, but over time we began seeing more and more problems with this setup. In this article we'll take a look at what we did to help solve these problems for both GitLab.com and self-hosted GitLab instances.

For example, the database was under constant pressure, with CPU utilization hovering around 70 percent almost all the time. Not because we used all available resources in the best way possible, but because we were bombarding the server with too many (badly optimized) queries. We realized we needed a better setup that would allow us to balance the load and make GitLab.com more resilient to any problems that may occur on the primary database server.

When tackling these problems using PostgreSQL there are essentially four techniques you can apply:

Optimize your application code so the queries are more efficient (and ideally use fewer resources).</li>

Use a connection pooler to reduce the number of database connections (and associated resources) necessary.</li>

Balance the load across multiple database servers.</li>

Shard your database.</li> </ol>
Optimizing the application code is something we have been working on actively for the past two years, but it's not a final solution. Even if you improve performance, when traffic also increases you may still need to apply the other two techniques. For the sake of this article we'll skip over this particular subject and instead focus on the other techniques.
Connection pooling</h2>
In PostgreSQL a connection is handled by starting an OS process which in turn needs a number of resources. The more connections (and thus processes), the more resources your database will use. PostgreSQL also enforces a maximum number of connections as defined in the max_connections</a> setting. Once you hit this limit PostgreSQL will reject new connections. Such a setup can be illustrated using the following diagram:

Here our clients connect directly to PostgreSQL, thus requiring one connection per client.
By pooling connections we can have multiple client-side connections reuse PostgreSQL connections. For example, without pooling we'd need 100 PostgreSQL connections to handle 100 client connections; with connection pooling we may only need 10 or so PostgreSQL connections depending on our configuration. This means our connection diagram will instead look something like the following:

Here we show an example where four clients connect to pgbouncer but instead of using four PostgreSQL connections we only need two of them.
For PostgreSQL there are two connection poolers that are most commonly used:

pgbouncer</a></li>
pgpool-II</a></li> </ul>
pgpool is a bit special because it does much more than just connection pooling: it has a built-in query caching mechanism, can balance load across multiple databases, manage replication, and more.
On the other hand pgbouncer is much simpler: all it does is connection pooling.
Database load balancing</h2>
Load balancing on the database level is typically done by making use of PostgreSQL's "hot standby</a>" feature. A hot-standby is a PostgreSQL replica that allows you to run read-only SQL queries, contrary to a regular standby that does not allow any SQL queries to be executed. To balance load you'd set up one or more hot-standby servers and somehow balance read-only queries across these hosts while sending all other operations to the primary. Scaling such a setup is fairly easy: simply add more hot-standby servers (if necessary) as your read-only traffic increases.
Another benefit of this approach is having a more resilient database cluster. Web requests that only use a secondary can continue to operate even if the primary server is experiencing issues; though of course you may still run into errors should those requests end up using the primary.
This approach however can be quite difficult to implement. For example, explicit transactions must be executed on the primary since they may contain writes. Furthermore, after a write we want to continue using the primary for a little while because the changes may not yet be available on the hot-standby servers when using asynchronous replication.
Sharding</h2>
Sharding is the act of horizontally partitioning your data. This means that data resides on specific servers and is retrieved using a shard key. For example, you may partition data per project and use the project ID as the shard key. Sharding a database is interesting when you have a very high write load (as there's no other easy way of balancing writes other than perhaps a multi-master setup), or when you have a lot of data and you can no longer store it in a conventional manner (e.g. you simply can't fit it all on a single disk).
Unfortunately the process of setting up a sharded database is a massive undertaking, even when using software such as Citus</a>. Not only do you need to set up the infrastructure (which varies in complexity depending on whether you run it yourself or use a hosted solution), but you also need to adjust large portions of your application to support sharding.
Cases against sharding</h3>
On GitLab.com the write load is typically very low, with most of the database queries being read-only queries. In very exceptional cases we may spike to 1500 tuple writes per second, but most of the time we barely make it past 200 tuple writes per second. On the other hand we can easily read up to 10 million tuples per second on any given secondary.
Storage-wise, we also don't use that much data: only about 800 GB. A large portion of this data is data that is being migrated in the background. Once those migrations are done we expect our database to shrink in size quite a bit.
Then there's the amount of work required to adjust the application so all queries use the right shard keys. While quite a few of our queries usually include a project ID which we could use as a shard key, there are also many queries where this isn't the case. Sharding would also affect the process of contributing changes to GitLab as every contributor would now have to make sure a shard key is present in their queries.
Finally, there is the infrastructure that's necessary to make all of this work. Servers have to be set up, monitoring has to be added, engineers have to be trained so they are familiar with this new setup, the list goes on. While hosted solutions may remove the need for managing your own servers it doesn't solve all problems. Engineers still have to be trained and (most likely very expensive) bills have to be paid. At GitLab we also highly prefer to ship the tools we need so the community can make use of them. This means that if we were going to shard the database we'd have to ship it (or at least parts of it) in our Omnibus packages. The only way you can make sure something you ship works is by running it yourself, meaning we wouldn't be able to use a hosted solution.
Ultimately we decided against sharding the database because we felt it was an expensive, time-consuming, and complex solution to a problem we do not have.
Connection pooling for GitLab</h2>
For connection pooling we had two main requirements:

It has to work well (obviously).</li>
It has to be easy to ship in our Omnibus packages so our users can also take advantage of the connection pooler.</li> </ol>
Reviewing the two solutions (pgpool and pgbouncer) was done in two steps:

Perform various technical tests (does it work, how easy is it to configure, etc).</li>
Find out what the experiences are of other users of the solution, what problems they ran into and how they dealt with them, etc.</li> </ol>
pgpool was the first solution we looked into, mostly because it seemed quite attractive based on all the features it offered. Some of the data from our tests can be found in this</a> comment.
Ultimately we decided against using pgpool based on a number of factors. For example, pgpool does not support sticky connections. This is problematic when performing a write and (trying to) display the results right away. Imagine creating an issue and being redirected to the page, only to run into an HTTP 404 error because the server used for any read-only queries did not yet have the data. One way to work around this would be to use synchronous replication, but this brings many other problems to the table; problems we prefer to avoid.
Another problem is that pgpool's load balancing logic is decoupled from your application and operates by parsing SQL queries and sending them to the right server. Because this happens outside of your application you have very little control over which query runs where. This may actually be beneficial to some because you don't need additional application logic, but it also prevents you from adjusting the routing logic if necessary.
Configuring pgpool also proved quite difficult due to the sheer number of configuration options. Perhaps the final nail in the coffin was the feedback we got on pgpool from those having used it in the past. The feedback we received regarding pgpool was usually negative, though not very detailed in most cases. While most of the complaints appeared to be related to earlier versions of pgpool it still made us doubt if using it was the right choice.
The feedback combined with the issues described above ultimately led to us deciding against using pgpool and using pgbouncer instead. We performed a similar set of tests with pgbouncer and were very satisfied with it. It's fairly easy to configure (and doesn't have that much that needs configuring in the first place), relatively easy to ship, focuses only on connection pooling (and does it really well), and had very little (if any) noticeable overhead. Perhaps my only complaint would be that the pgbouncer website can be a little bit hard to navigate.
Using pgbouncer we were able to drop the number of active PostgreSQL connections from a few hundred to only 10-20 by using transaction pooling. We opted for using transaction pooling since Rails database connections are persistent. In such a setup, using session pooling would prevent us from being able to reduce the number of PostgreSQL connections, thus brining few (if any) benefits. By using transaction pooling we were able to drop PostgreSQL's max_connections</code> setting from 3000 (the reason for this particular value was never really clear) to 300. pgbouncer is configured in such a way that even at peak capacity we will only need 200 connections; giving us some room for additional connections such as psql</code> consoles and maintenance tasks.
A side effect of using transaction pooling is that you cannot use prepared statements, as the PREPARE</code> and EXECUTE</code> commands may end up running in different connections; producing errors as a result. Fortunately we did not measure any increase in response timings when disabling prepared statements, but we did measure a reduction of roughly 20 GB in memory usage on our database servers.
To ensure both web requests and background jobs have connections available we set up two separate pools: one pool of 150 connections for background processing, and a pool of 50 connections for web requests. For web requests we rarely need more than 20 connections, but for background processing we can easily spike to a 100 connections simply due to the large number of background processes running on GitLab.com.
Today we ship pgbouncer as part of GitLab EE's High Availability package. For more information you can refer to"Omnibus GitLab PostgreSQL High Availability."</a>
Database load balancing for GitLab</h2> With pgpool and its load balancing feature out of the picture we needed something else to spread load across multiple hot-standby servers. For (but not limited to) Rails applications there is a library called Makara</a> which implements load balancing logic and includes a default implementation for ActiveRecord. Makara however has some problems that were a deal-breaker for us. For example, its support for sticky connections is very limited: when you perform a write the connection will stick to the primary using a cookie, with a fixed TTL. This means that if replication lag is greater than the TTL you may still end up running a query on a host that doesn't have the data you need. Makara also requires you to configure quite a lot, such as all the database hosts and their roles, with no service discovery mechanism (our current solution does not yet support this either, though it's planned for the near future). Makara also does not appear to be thread-safe</a>, which is problematic since Sidekiq (the background processing system we use) is multi-threaded. Finally, we wanted to have control over the load balancing logic as much as possible. Besides Makara there's also Octopus</a> which has some load balancing mechanisms built in. Octopus however is geared towards database sharding and not just balancing of read-only queries. As a result we did not consider using Octopus. Ultimately this led to us building our own solution directly into GitLab EE. The merge request adding the initial implementation can be found here</a>, though some changes, improvements, and fixes were applied later on. Our solution essentially works by replacing ActiveRecord::Base.connection</code> with a proxy object that handles routing of queries. This ensures we can load balance as many queries as possible, even queries that don't originate directly from our own code. This proxy object in turn determines what host a query is sent to based on the methods called, removing the need for parsing SQL queries. Sticky connections</h3> Sticky connections are supported by storing a pointer to the current PostgreSQL WAL position the moment a write is performed. This pointer is then stored in Redis for a short duration at the end of a request. Each user is given their own key so that the actions of one user won't lead to all other users being affected. In the next request we get the pointer and compare this with all the secondaries. If all secondaries have a WAL pointer that exceeds our pointer we know they are in sync and we can safely use a secondary for our read-only queries. If one or more secondaries are not yet in sync we will continue using the primary until they are in sync. If no write is performed for 30 seconds and all the secondaries are still not in sync we'll revert to using the secondaries in order to prevent somebody from ending up running queries on the primary forever. Checking if a secondary has caught up is quite simple and is implemented in Gitlab::Database::LoadBalancing::Host#caught_up?</code> as follows: def caught_up?(location) string = connection.quote(location) query = "SELECT NOT pg_is_in_recovery() OR " \ "pg_xlog_location_diff(pg_last_xlog_replay_location(), #{string}) >= 0 AS result" row = connection.select_all(query).first row && row['result'] == 't' ensure release_connection end </code></pre> Most of the code here is standard Rails code to run raw queries and grab the results. The most interesting part is the query itself, which is as follows: SELECT NOT pg_is_in_recovery() OR pg_xlog_location_diff(pg_last_xlog_replay_location(), WAL-POINTER) >= 0 AS result" </code></pre> Here WAL-POINTER</code> is the WAL pointer as returned by the PostgreSQL function pg_current_xlog_insert_location()</code>, which is executed on the primary. In the above code snippet the pointer is passed as an argument, which is then quoted/escaped and passed to the query. Using the function pg_last_xlog_replay_location()</code> we can get the WAL pointer of a secondary, which we can then compare to our primary pointer using pg_xlog_location_diff()</code>. If the result is greater than 0 we know the secondary is in sync. The check NOT pg_is_in_recovery()</code> is added to ensure the query won't fail when a secondary that we're checking was just promoted to a primary and our GitLab process is not yet aware of this. In such a case we simply return true</code> since the primary is always in sync with itself. Background processing</h3> Our background processing code always uses the primary since most of the work performed in the background consists of writes. Furthermore we can't reliably use a hot-standby as we have no way of knowing whether a job should use the primary or not as many jobs are not directly tied into a user. Connection errors</h3> To deal with connection errors our load balancer will not use a secondary if it is deemed to be offline, plus connection errors on any host (including the primary) will result in the load balancer retrying the operation a few times. This ensures that we don't immediately display an error page in the event of a hiccup or a database failover. While we also deal with hot standby conflicts</a> on the load balancer level we ended up enabling hot_standby_feedback</code> on our secondaries as doing so solved all hot-standby conflicts without having any negative impact on table bloat. The procedure we use is quite simple: for a secondary we'll retry a few times with no delay in between. For a primary we'll retry the operation a few times using an exponential backoff. For more information you can refer to the source code in GitLab EE: https://gitlab.com/gitlab-org/gitlab-ee/tree/master/lib/gitlab/database/load_balancing.rb</a></li> https://gitlab.com/gitlab-org/gitlab-ee/tree/master/lib/gitlab/database/load_balancing</a></li> </ul> Database load balancing was first introduced in GitLab 9.0 and only supports PostgreSQL. More information can be found in the 9.0 release post</a> and the documentation</a>. Crunchy Data</h2> In parallel to working on implementing connection pooling and load balancing we were working with Crunchy Data</a>. Until very recently I was the only database specialist</a> which meant I had a lot of work on my plate. Furthermore my knowledge of PostgreSQL internals and its wide range of settings is limited (or at least was at the time), meaning there's only so much I could do. Because of this we hired Crunchy to help us out with identifying problems, investigating slow queries, proposing schema optimisations, optimising PostgreSQL settings, and much more. For the duration of this cooperation most work was performed in confidential issues so we could share private data such as log files. With the cooperation coming to an end we have removed sensitive information from some of these issues and opened them up to the public. The primary issue was gitlab-com/infrastructure#1448</a>, which in turn led to many separate issues being created and resolved. The benefit of this cooperation was immense as it helped us identify and solve many problems, something that would have taken me months to identify and solve if I had to do this all by myself. Fortunately we recently managed to hire our second database specialist</a> and we hope to grow the team more in the coming months. Combining connection pooling and database load balancing</h2> Combining connection pooling and database load balancing allowed us to drastically reduce the number of resources necessary to run our database cluster as well as spread load across our hot-standby servers. For example, instead of our primary having a near constant CPU utilisation of 70 percent today it usually hovers between 10 percent and 20 percent, while our two hot-standby servers hover around 20 percent most of the time: Here db3.cluster.gitlab.com</code> is our primary while the other two hosts are our secondaries. Other load-related factors such as load averages, disk usage, and memory usage were also drastically improved. For example, instead of the primary having a load average of around 20 it barely goes above an average of 10: During the busiest hours our secondaries serve around 12 000 transactions per second (roughly 740 000 per minute), while the primary serves around 6 000 transactions per second (roughly 340 000 per minute): Unfortunately we don't have any data on the transaction rates prior to deploying pgbouncer and our database load balancer. An up-to-date overview of our PostgreSQL statistics can be found at our public Grafana dashboard</a>. Some of the settings we have set for pgbouncer are as follows: Setting</th> Value</th> </tr> </thead> default_pool_size</td> 100</td> </tr> reserve_pool_size</td> 5</td> </tr> reserve_pool_timeout</td> 3</td> </tr> max_client_conn</td> 2048</td> </tr> pool_mode</td> transaction</td> </tr> server_idle_timeout</td> 30</td> </tr> </tbody> </table> With that all said there is still some work left to be done such as: implementing service discovery (#2042</a>), improving how we check if a secondary is available (#2866</a>), and ignoring secondaries that are too far behind the primary (#2197</a>). It's worth mentioning that we currently do not have any plans of turning our load balancing solution into a standalone library that you can use outside of GitLab, instead our focus is on providing a solid load balancing solution for GitLab EE. If this has gotten you interested and you enjoy working with databases, improving application performance, and adding database-related features to GitLab (such as service discovery</a>) you should definitely check out the job opening</a> and the database specialist handbook entry</a> for more information. GitLab Patch Release: 9.5.6 2017-09-28T00:00:00+00:00 Today we are releasing version 9.5.6 for GitLab Community Edition (CE) and Enterprise Edition (EE). This version resolves a number of regressions and bugs in this month's 9.5.5 release</a> and prior versions. It includes the following fixes: CE/EE Fix MR widget with ready to merge buttons/controls at mobile breakpoint (!14242</a>)</li> CE/EE Fix MR widget with external CI services/integrations (!13977</a>)</li> CE/EE Update "x/x discussions resolved" checkmark icon to be green when all discussions resolved (!14255</a>)</li> CE/EE Upgrade Nokogiri because of CVE-2017-9050 (!14427</a>)</li> CE/EE Handle error when fetching ref for MR with deleted source branch (!14550</a>)</li> EE Show group tab if member lock is enabled (!3025</a>)</li> Omnibus CE/EE Bump ruby version to 2.3.5 for security fixes (!1930</a>)</li> Omnibus CE/EE Resolve "update_databases assumes a populated databases.json" (!1924</a>)</li> Omnibus CE/EE Resolve "restart repmgrd after node register" (!1928</a>)</li> Omnibus CE/EE Resolve "repmgr unregister doesn't accept –node" (!1927</a>)</li> </ul> Upgrade barometer</h2> This version does not include any new migrations, and should not require any downtime. Please be aware that by default the Omnibus packages will stop, run migrations, and start again, no matter how “big” or “small” the upgrade is. This behavior can be changed by adding a /etc/gitlab/skip-auto-migrations</code> file</a>. Updating</h2> To update, check out our update page</a>. Enterprise Edition</h2> Interested in GitLab Enterprise Edition? Check out the features exclusive to EE</a>. Access to GitLab Enterprise Edition is granted by a subscription</a>. No time to upgrade GitLab yourself? Subscribers receive upgrade and installation services. GitLab Patch Release: 10.0.2 2017-09-27T00:00:00+00:00 Today we are releasing version 10.0.2 for GitLab Community Edition (CE) and Enterprise Edition (EE). This version resolves a number of regressions and bugs in this month's 10.0.1 release</a> and prior versions. It includes the following fixes: CE/EE: Resize comment form after note submit and discard it (!14401</a>)</li> CE/EE: Fix rendering double note issue (!14429</a>)</li> CE/EE: Ensure that Blob.raw returns always a valid blob object (!14412</a>)</li> CE/EE: Remove navbar gradient (!14457</a>)</li> CE/EE: Re-allow name</code> attribute on user-provided anchor HTML (!14452</a>)</li> CE/EE: Fix checks failing with 'undefined method run_command</code> in rake gitlab:check</code> (!14469</a>)</li> CE/EE: Fix PG::UniqueViolation external_email migration (!14474</a>)</li> CE/EE: Prevent URL concatenation for avatars (!14437</a>)</li> CE/EE: IssueNotes: Switch back to Write pane when note cancel or submit (!14402</a>)</li> CE/EE: Fix a dash being rendered in the note's access role (!14450</a>)</li> CE/EE: Make locked setting of Runner to not affect jobs scheduling (!14483</a>)</li> CE/EE: Avoid wrapping users in dropdown (!14410</a>)</li> EE: Fix delta size check to handle commit or nil objects (!2976</a>)</li> EE: Fix 500 error due to promote group Web hooks (!2972</a>)</li> EE: Add notes on authentication to PG HA document (!2851</a>)</li> EE: Send valid project path as name for Jira dev panel (!3005</a>)</li> EE: Remove duplicate group boards doc (!2991</a>)</li> EE: Point to embedded Consul (!2980</a>)</li> EE: Fix internal link in docs(!3003</a>)</li> Omnibus: Fix an issue where enabling a GitLab Geo role would also disable all default services (!1963</a>)</li> Omnibus: Add option to configure redis snapshot frequency (!1964</a>)</li> Omnibus: Update pg-upgrade output to be more clear when a bundled PostgreSQL is not in use (!1962</a>)</li> Omnibus: Reload consul on config change instead of restarting (!1966</a>)</li> </ul> Upgrade barometer</h2> This version does not include any new migrations, and should not require any downtime. Please be aware that by default the Omnibus packages will stop, run migrations, and start again, no matter how “big” or “small” the upgrade is. This behavior can be changed by adding a /etc/gitlab/skip-auto-migrations</code> file</a>. Updating</h2> To update, check out our update page</a>. Enterprise Edition</h2> Interested in GitLab Enterprise Edition? Check out the features exclusive to EE</a>. Access to GitLab Enterprise Edition is granted by a subscription</a>. No time to upgrade GitLab yourself? Subscribers receive upgrade and installation services. GitLab Continuous Integration named a Leader in the Forrester Wave™ 2017-09-27T00:00:00+00:00 Today Forrester has evaluated GitLab as a Leader in Continuous Integration in The Forrester Wave™: Continuous Integration Tools, Q3 2017 report. As the report states, “GitLab delivers ease of use, scalability, integration, and innovation.” We see CI/CD as a critical component of our offering, making it quicker and easier to deliver value through automation. The progress we’ve made in developing a CI/CD solution that’s seamlessly integrated is thanks largely to the participation and feedback of our community and customers. We’re excited about what the future holds for GitLab CI/CD</a>, so it’s great to know that others see its value. Furthermore Forrester states, “GitLab’s vision is to serve enterprise-scale, integrated software development teams that want to spend more time writing code and less time maintaining their tool chain.” We’ve said it before (we say it a lot, actually): CI/CD is the cornerstone of modern software development. Delivering what customers want, faster, requires a modernized software development lifecycle that saves time, effort, and cost. Automation is key to this more efficient release cycle, which is why we’ve spent a lot of time developing our built-in test and release automation, so developers can spend less time stringing together their tool chain, and more on innovation and integrating customer feedback. We’re gratified to have been named a leader in CI by Forrester in their report on Continuous Integration Tools for Q3, 2017, which evaluated 10 vendors across 25 different criteria. We had some key areas of success: Top score in the Current Offering category</h2> GitLab received the highest score in Forrester’s Current Offering evaluation, which we believe confirms that our product is hard to beat when it comes to ease of installation, configuring builds, platform support, analytics, an intuitive UI, container support and more. This all adds up to a more seamless development process which facilitates DevOps practices and collaboration. Highest score possible in the Strategy category</h2> Forrester assessed our product strategy, market approach, training, consulting and support, giving us a 5.0, the highest possible score, for all three criteria. This is huge validation for our CI/CD vision, shaped with the help of our customers, community, and their contributions. Just the beginning</h2> A significant part of that vision, Auto DevOps, just shipped with our 10.0 release</a>. Auto DevOps is a hands-free DevOps experience that automatically configures your build, test, code quality assurance, review apps, deployment, and monitoring in a single environment. Our out-of-the-box templates allow you to set up an end-to-end DevOps lifecycle at the push of a button. Auto DevOps is just the first step on a journey to building a complete and highly automated DevOps tool chain. Building on strong momentum, we’re working on a number of new features to better leverage our CI/CD tools in a more automated way for both operations as well as development teams. With the full automation we are planning, you will not only be able to deliver a new, functioning application from a developer’s machine to a production environment in minutes, but also monitor all production environments in one dashboard. Faster time to value</h3> Reducing the time teams spend on manual tasks means you can focus on what matters: building great products that your users want. GitLab CI/CD facilitates a more iterative approach to software development, by automatically testing new code, helping you to catch potential problems earlier in the release cycle, taking the anxiety out of deployments and reducing the cost of fixing bugs. This puts you in a position to release more often, shortening the feedback loop and helping you to release features and improvements that your customers are asking for. Collaboration without dependency</h3> The fact that GitLab CI/CD is fully integrated makes all the difference. Less context-switching means key information doesn’t slip through the cracks like it can when teams work in different tools, with no single source of truth. Working in an integrated tool fosters collaboration by improving visibility, keeping everyone on the same page and making it easy for different teams to participate and comment. Features like review apps – which automatically spin up a dynamic environment for merge requests, so you can preview changes right away – make it even easier to share improvements with stakeholders and gather feedback. Being named a Leader in this Wave evaluation, and receiving the Top Score in Current Offering, as well as the highest score possible for Strategy, is validation for us of the strength of today’s GitLab CI offering as well as our vision for the future. View the full report</a> to learn more about GitLab Continuous Integration. GitLab Patch Release: 10.0.1 2017-09-24T00:00:00+00:00 Today we are releasing version 10.0.1 for GitLab Community Edition (CE) and Enterprise Edition (EE). This version resolves a migration issue in this month's 10.0.0 release</a> that affected instances with inactive LDAP accounts</a>. Other than that, this update contains no changes to functionality from 10.0.0. It includes the following fix: CE/EE Fix duplicate key errors in PostDeployMigrateUserExternalMailData migration. (!14460</a>)</li> </ul> Thanks to Ian Yang for suggesting a solution. Upgrade barometer</h2> This version includes a revision to a single migration, but does not require any downtime. If you have already successfully upgraded to 10.0.0, this upgrade will not affect you. Please be aware that by default the Omnibus packages will stop, run migrations, and start again, no matter how “big” or “small” the upgrade is. This behavior can be changed by adding a /etc/gitlab/skip-auto-migrations</code> file</a>. Updating</h2> To update, check out our update page</a>. Enterprise Edition</h2> Interested in GitLab Enterprise Edition? Check out the features exclusive to EE</a>. Access to GitLab Enterprise Edition is granted by a subscription</a>. No time to upgrade GitLab yourself? Subscribers receive upgrade and installation services. GitLab 10.0 Released with Auto DevOps and Group Issue Boards 2017-09-22T00:00:00+00:00 From the formulation of an idea to executing and monitoring it in production, DevOps establishes a culture and environment where developing, testing, and releasing software can happen quickly, frequently, and more reliably. GitLab 10.0 delivers a hands-free DevOps environment with the introduction of Auto DevOps</a>, allowing your team to easily configure and adopt modern development practices in your workflow. Not only that, there's new navigation</a> and a new way of collaborating across groups</a>. With every monthly release of GitLab, we introduce new capabilities and improve our existing features. GitLab 10.0 is no exception and includes numerous new additions, such as the ability to automatically resolve outdated merge request discussions</a>, improvements to subgroups</a>, and an API for Wiki</a> thanks to a contribution from our open source community. GitLab's powerful issue management capabilities keep getting better with every release. Filtering and searching issues across groups has been vastly improved</a>, our updated UX makes moving issues easier to discover</a> and can be automated through quick action commands</a>. GitLab Enterprise Edition Premium customers using JIRA can now see commits and branches</a> in JIRA's development panel. Security and performance continues to improve. Administrators can now restrict SSH</a> access through technology and key length. LDAP Group Sync can be automated</a> through our API and can now lock down External Users</a> at point of login as well. Performance</a> continues to get faster, improving page loading speeds, the speed of creating projects and performing commits, and reduced memory usage.

Watch</h2>
We'll be streaming on YouTube</a>. You can watch Sid</a> and Dmitriy</a>'s keynotes, our Santorini trip, GitLabber-led user generated content (UGC) sessions</a>, the 10.1 release, and our final party can all be viewed in one place.</p> </iframe>

Cross-Site Scripting (XSS) vulnerability in search bar</h2>
Josh Unger</a> reported a Cross-Site Scripting (XSS) vulnerability in the issue search bar. Usernames were not being properly HTML escaped inside the author filter would could allow arbitrary script execution. #38267</a></p>

Libxml2 update</h2>
The version of libxml2 included with GitLab Omnibus CE+EE packages has been updated to 2.9.6 to patch several security vulnerabilities. XMLSoft</a></p>

Updating</h2>
To update, check out our update page</a>.</p>

Enterprise Edition</h2>
Interested in GitLab Enterprise Edition? Check out the features exclusive to EE</a>.</p>
Access to GitLab Enterprise Edition is included with a subscription</a>. No time to upgrade GitLab yourself? Subscribers receive upgrade and installation services.</p>

About the guest author</h3>
Paul Hibbitts has been an interaction design practitioner and educator for over 20 years, and has recently ventured into the world of open source software development thanks to the amazing Grav CMS.</p>
Cover image</a> by chuttersnap</a> on unsplash</p>

Enterprise Edition</h2>
Interested in GitLab Enterprise Edition? Check out the features exclusive to EE</a>.</p>
Access to GitLab Enterprise Edition is granted by a subscription</a>. No time to upgrade GitLab yourself? Subscribers receive upgrade and installation services.</p>

Enterprise Edition</h2>
Interested in GitLab Enterprise Edition? Check out the features exclusive to EE</a>.</p>
Access to GitLab Enterprise Edition is granted by a subscription</a>. No time to upgrade GitLab yourself? Subscribers receive upgrade and installation services.</p>

Enterprise Edition</h2>
Interested in GitLab Enterprise Edition? Check out the features exclusive to EE</a>.</p>
Access to GitLab Enterprise Edition is granted by a subscription</a>. No time to upgrade GitLab yourself? Subscribers receive upgrade and installation services.</p>

GitLab Patch Release: 9.5.6

Enterprise Edition</h2>
Interested in GitLab Enterprise Edition? Check out the features exclusive to EE</a>.</p>
Access to GitLab Enterprise Edition is granted by a subscription</a>. No time to upgrade GitLab yourself? Subscribers receive upgrade and installation services.</p>

GitLab Patch Release: 10.0.2

Enterprise Edition</h2>
Interested in GitLab Enterprise Edition? Check out the features exclusive to EE</a>.</p>
Access to GitLab Enterprise Edition is granted by a subscription</a>. No time to upgrade GitLab yourself? Subscribers receive upgrade and installation services.</p>

GitLab Continuous Integration named a Leader in the Forrester Wave™

GitLab Patch Release: 10.0.1

Enterprise Edition</h2>
Interested in GitLab Enterprise Edition? Check out the features exclusive to EE</a>.</p>
Access to GitLab Enterprise Edition is granted by a subscription</a>. No time to upgrade GitLab yourself? Subscribers receive upgrade and installation services.</p>

GitLab 10.0 Released with Auto DevOps and Group Issue Boards

GitLab

GitLab 10.1 Released with Image Discussions and Rejection of Unsigned Commits

We're coming to you live from Crete, at the GitLab Summit!

GitLab 10.0.4, 9.5.9, and 9.4.7 Released

How to spot development issues and fix them fast

Enabling an open and collaborative course environment with GitLab and the Grav CMS

Complete DevOps is DevOps reimagined. Here's what that looks like

Announcing $20 million in Series C round funding led by GV to complete DevOps

GitLab Patch Release: 10.0.3

Beyond CI/CD: GitLab's DevOps vision

GitLab Patch Release: 9.5.8

GitLab Patch Release: 9.5.7

Scaling the GitLab database

GitLab

GitLab 10.1 Released with Image Discussions and Rejection of Unsigned Commits

We're coming to you live from Crete, at the GitLab Summit!

GitLab 10.0.4, 9.5.9, and 9.4.7 Released

Versions affected</h3> Cross-Site Scripting (XSS) vulnerability in markdown:</p> GitLab CE+EE 2.8.0-9.4.6, 9.5.0-9.5.8, 10.0.0-10.0.3</li> </ul> Cross-Site Scripting (XSS) vulnerability in search bar</p>

How to spot development issues and fix them fast

Track developer stress and coping strategies</h2> Start with getting to know the team behind the work – learn how to track their stress levels accurately.</strong></p>

Enabling an open and collaborative course environment with GitLab and the Grav CMS

Complete DevOps is DevOps reimagined. Here's what that looks like

Announcing $20 million in Series C round funding led by GV to complete DevOps

GitLab Patch Release: 10.0.3

Upgrade barometer</h2> This version does not include any new migrations, and should not require any downtime.</p>

Beyond CI/CD: GitLab's DevOps vision

The DevOps tool chain</h3> To offer some context: DevOps is hard to define, because everybody defines it slightly differently. Sometimes DevOps is defined as the intersection of development, operations, and quality assurance.</p> </p>

GitLab Patch Release: 9.5.8

Upgrade barometer</h2> This version does not include any new migrations, and should not require any downtime.</p>

GitLab Patch Release: 9.5.7

Upgrade barometer</h2> This version does not include any new migrations, and should not require any downtime.</p>

Scaling the GitLab database

GitLab Patch Release: 9.5.6

Upgrade barometer</h2> This version does not include any new migrations, and should not require any downtime.</p>

GitLab Patch Release: 10.0.2

Upgrade barometer</h2> This version does not include any new migrations, and should not require any downtime.</p>

GitLab Continuous Integration named a Leader in the Forrester Wave™

GitLab Patch Release: 10.0.1

Upgrade barometer</h2> This version includes a revision to a single migration, but does not require any downtime. If you have already successfully upgraded to 10.0.0, this upgrade will not affect you.</p>

GitLab 10.0 Released with Auto DevOps and Group Issue Boards