The performance of quota usage sync in cinder driver should be enchanced

Bug #1583965 reported by Chaoyi Huang
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Fix Committed
Goutham Pratapa

Bug Description

Currently the quota usage is sync-ed through list of all resources of these project and then sum. This handling's performance is not good, especially if multiple quota-usage task executed concurrently, will heavily impact the performance of Cinder for many queries in parallel

                volumes = self.cinder.volumes.list(search_opts=opts)
                snapshots = self.cinder.volume_snapshots.list(search_opts=opts)
                backups = self.cinder.backups.list(search_opts=opts)

                usages['gigabytes'] = sum([int(v.size) for v in volumes])
                usages['volumes'] = len(volumes)
                usages['snapshots'] = len(snapshots)
                usages['backups'] = len(backups)

After Cinder client support limit calling to get the usage directly, the code here should be improved.

Revision history for this message
Chaoyi Huang (joehuang) wrote :

Dimitri has address the query issue in Cinder, it's time to optimize the query in kingbird

Changed in kingbird:
assignee: nobody → Goutham Pratapa (pratapagoutham)
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix proposed to kingbird (master)

Fix proposed to branch: master

Changed in kingbird:
status: New → In Progress
Changed in kingbird:
status: In Progress → Fix Committed
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Change abandoned on kingbird (master)

Change abandoned by Goutham Pratapa (<email address hidden>) on branch: master

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers