Closed Bug 1598091 Opened 5 years ago Closed 5 years ago

Is an extra, or missing, index the cause of the extra database writes?

Categories

(Tree Management :: Treeherder: Infrastructure, defect)

Product:
Tree Management
Component:
Treeherder: Infrastructure
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

(Not tracked)

Status:
RESOLVED INVALID

People

(Reporter: ekyle, Assigned: ekyle)

References

(Blocks 1 open bug)

Details

In theory, a query that must sort data can cause extra writes if the sort table is too big to fit in memory. Poor choice of indexes can cause this.

  • Did the indexes de-synchronize between prod and stage?
  • The volume of data that is sorted can cause the different behaviour
Assignee: nobody → klahnakoski

Prod and stage are the same in terms of indexes

Blocks: 1599095

Last week, during the outage, I confirmed that the database schema are identical. I removed one index in staging to see if there was a difference: There was no change, so the corresponding index was not removed from production.

I was late to realize the amount of data in both stage and prod were almost identical.

Status: NEW → RESOLVED
Closed: 5 years ago
Resolution: --- → INVALID
You need to log in before you can comment on or make changes to this bug.