Data Center

2 statements.

Author	Tags	Primary Session	Secondary Sessions	Position Statement
Roan Kattouw	Architecture, Contributors, Data Center	Evolving the MediaWiki Architecture	Advancing the Contributor Experience	Users should not be punished for logging in WMF wikis are slower for logged-in users than for anonymous users, which is unhelpful for trying to get users to contribute. This is a long standing problem that's hard to solve, but we should have a vision for how we're going to solve it. WMF has caching data centers in strategic locations around the world (Amsterdam, San Francisco and soon Singapore), which make the wikis faster for users who are not near the primary data center (in Virginia) but are near a caching location. However, this only benefits anonymous users. For logged-in users, every page view contains their user name and other user-specific information in the personal tools area, so logged-in page views are considered uncacheable and are always routed to the primary data center. This means that if a new user browses the site for a while, then creates an account because they want to contribute (or makes an anonymous edit), the wiki suddenly becomes slower for them. All users are affected, because uncached requests are slower to serve than cached ones, but users outside North/South America are affected the most, because their traffic now has to cross an ocean that it didn't have to cross before. It's not nice that a new user's 'reward' for creating an account is a slower experience, but it's especially not nice that users in emerging communities are affected the most. If we want to encourage readers to become contributors, slowing the site down as soon as someone contributes is not very helpful. Some requests will always have to go to the primary data center, such as POST requests saving an edit, and those are always going to be slower for users outside North America. But for logged-in page views this isn't fundamentally necessary, and serving them from the edge caches would speed up the site for logged-in users and reduce the load on the app servers. There are different ways that this could be done, each with their own obstacles. For example, a single-page application for MediaWiki could use a content service to retrieve only the new page's contents when navigating, but this would require modifying or rewriting a lot of code in MediaWiki; ESI could be used to have Varnish inject cached page contents into a user-specific chrome, but that would require using advanced and partly unproven Varnish features. In both cases, we'd have to reimplement certain rendering preferences using CSS or a post-processing step. It's far from trivial, but let's start talking seriously about how we can address this problem.
Giuseppe Lavagetto	Analytics, Data Center, Infrastructure, Job Queue, Multi-Datacenter, Performance	Evolving the MediaWiki Architecture		The future of the MediaWiki infrastructure at the wikimedia foundation. MediaWiki is at the core of the infrastructure that serves all of the Wikimedia projects, and the current setup of MediaWiki in production poses various challenges: from the future of our current runtime (HHVM), to the to ability to serve MediaWiki from multiple DataCenters, to long standing issues as resource usage efficiency and flexibility. Here are some of the things we have to tackle in the future. Transition off of HHVM: Since the HHVM team has made it clear they're parting ways with full PHP compatibility, and that maintaining support for both HHVM and PHP in MediaWiki would be arduous, we need to make plans to move off of HHVM, back to PHP 7.x. This transition, while technically necessary, should not come at a cost for our users: page load times should not degrade. We can proceed by marking responses coming from either engine, collecting metrics and analyzing data. In order to achieve this, we should run the two runtimes in parallel on the same servers (which have plenty of capacity, given no MediaWiki cluster has an utilization over 40%), and we will then be able to programmatically route individual users or a percentage of traffic, or even specific wikis, to one or the other. The deadline for this transition is set for the end of 2018 (EOL of the last compatible version of HHVM), and planning and resources should be allocated to this goal. Multi-Datacenter support: We currently are using our datacenters in a active/passive setting, as far as MediaWiki is concerned. While this is ok in line of principle, this is a huge waste of resources and means we both have 50% of our servers doing nothing at all, and also limits our ability to expand the number of core datacenters we can use. Diverting the read load to secondary datacenters could also allow us to use caching in a less aggressive way when not needed. There is already a program underway to add first-class multi-DC support to MediaWiki, so we can focus on what specifically needs to be done in order to achieve this longtime goal: our final goal should be to serve reader's traffic from all datacenters, and to be able to switch the "master" datacenter in matter of minutes. Elasticity, resource usage efficiency At the moment , our infrastructure is plainly inadequate to react to sudden spikes of non-wiki content production and to changes that generate a lot of asynchrounous jobs, as a change of a popular template. The issues with the current jobqueue are widely known and publicized, but even the current transition to a new model won't solve the starvation of resources that result in a degraded user experience. Moreover, a single editor uploading videos via video2commons can easily overflow our media processing capacity for weeks. This happens because we allocate our resource statically (we have 4 vidoesclaers per datcenter, for example), we have an inefficient resource consumption, and reallocating servers requires time and effort. Modern applications stacks are elastic, meaning the operation of scaling up or down the capacity of a single cluster or functionality can be handled programmatically and/or manually whenever the need occurs, allowing the infrastructure to react to such changes. For economic and privacy/security reasons, Wikimedia doesn't make use of external cloud services, so the only way to achieve such flexibility is to build a serviceable infrastructure that can serve MediaWiki and any other project Wikimedia will support: the effort to do that is underway with the rollout of our Kubernetes-based IaaS in production. I think we should work, sooner than later, at moving the MediaWiki application stack (and maybe its semi-ephemeral caching) to the kubernetes platform. While the advantages of such an approach seem clear, it won't come without costs: specifically habits around code deployment, testing and configuration changes will need to be completely revisited and superseded by new approaches.