Difference between revisions of "Large Scale SIG/ScaleUp"

Revision as of 08:34, 16 December 2020

The third stage in the Scaling Journey is Scale Up.

As you monitor your cluster at scale, you will see that it hits scaling limits within one cluster. All hope is not lost, though! There are things you can put in place push back how much a single cluster can handle, before having to resort to setting up a more complex deployment configuration. This page aims to help answer those questions.

Once you are past that stage, you are ready to proceed to next stage of the Scaling Journey: Scale Out.

FAQ

Q: Cleaning up deleted entries in my database is a bit of a hassle. is there a tool I could use to help me with that?

A: The OSarchiver tool, developed by OVH, can help you there: see https://github.com/ovh/osarchiver/ . We are working on making it maintained upstream as part of the OSops tooling.

Q: How many compute nodes can a typical OpenStack cluster contain ?

A: Request may timeout when scheduling large number of instances in a single request (> 100) when cluster size grows beyond 1000 compute nodes

Q: RabbitMQ is clearly my bottleneck, is there a way to reconfigure that part to handle more scale?

Resources

A curated collection of scaling stories, as we collect them
Neutron backends performance comparison: https://thesaitech.wordpress.com/2019/02/15/a-comparative-study-of-openstack-networking-architectures/
Evaluation of internal messaging
Evaluation of databases
- https://www.openstack.org/summit/vancouver-2018/summit-schedule/events/21212/keystone-in-the-context-of-fogedge-massively-distributed-clouds
- https://beyondtheclouds.github.io/blog/openstack/cockroachdb/2018/06/04/evaluation-of-openstack-multi-region-keystone-deployments.html
Scaling Neutron: https://www.youtube.com/watch?v=5WL47L1P5kE (https://www.slideshare.net/moreirabelmiro/evolution-of-openstack-networking-at-cern)
Scaling Nova/Ironic: https://techblog.web.cern.ch/techblog/post/nova-ironic-at-scale/
Scheduling Performance: https://techblog.web.cern.ch/techblog/post/scheduling-optimizations/
Global scaling: https://www.openstack.org/summit/barcelona-2016/summit-schedule/events/15977/chasing-1000-nodes-scale

Other SIG work on that stage

Collecting scaling stories
- Submit scaling stories on https://etherpad.openstack.org/p/scaling-stories
- Curate them on Large_Scale_Scaling_Stories

@@ Line 14: / Line 14: @@
 '''Q: How many compute nodes can a typical OpenStack cluster contain ?'''
-A: ...
+A: Request may timeout when scheduling large number of instances in a single request (> 100) when cluster size grows beyond 1000 compute nodes
 '''Q: RabbitMQ is clearly my bottleneck, is there a way to reconfigure that part to handle more scale?'''