Project

General

Profile

Actions

Bug #7987

closed

rudder-upgrade takes too long

Added by Janos Mattyasovszky over 8 years ago. Updated over 8 years ago.

Status:
Released
Priority:
N/A
Category:
Packaging
Target version:
Severity:
UX impact:
User visibility:
Effort required:
Priority:
Name check:
Fix check:
Regression:

Description

The pgsql query

select count(*) from (select xpath('/entry/*[@fileFormat=<?>]',data) AS x from eventlog) as Y where array_upper(x, 1) > 0
takes too long if you have 300k+ entries.

Server: SLES11 SP3 // postgresql91-server-9.1.18-0.3.1

This is also heavily impacted by the fact, that the script /opt/rudder/bin/rudder-upgrade, that contains this statement, is run twice on a single-server installation:

A 3.0.11->3.0.13 update took:

  rpm rudder-inventory-ldap: 61 minutes
  rpm rudder-webapp:         61 minutes

This could been solved if you would:
  • Save each type of update that is not required, because it has been checked during any previous update.
  • Execute this kind of maintenance by the jetty in an async job, not during the RPM installation.

Subtasks 2 (0 open2 closed)

Bug #8001: create an index for detection of eventlog versionReleasedFrançois ARMAND2016-02-29Actions
Bug #8002: Merge error in branch 3.1 for fileFormat detectionReleasedFrançois ARMAND2016-02-29Actions
Actions

Also available in: Atom PDF