Bug #19973: Performance problem - ldap high IO read and continuous re-generation - Rudder - Issue Tracker

Actions

Copy link

Bug #19973

closed

Performance problem - ldap high IO read and continuous re-generation

Added by Alexis Mousset about 3 years ago. Updated over 1 year ago.

Status:

Rejected

Priority:

N/A

Assignee:

Nicolas CHARLES

Category:

Web - Config management

Target version:

old 6.2 issues to relocate

Pull Request:

Severity:

Major - prevents use of part of Rudder | no simple workaround

UX impact:

User visibility:

Operational - other Techniques | Rudder settings | Plugins

Effort required:

Priority:

Name check:

To do

Fix check:

To do

Regression:

Description

We saw this problem recently on several instances. The symptoms are:

Continuous reg-generation, and everything else (UI/API/etc.) is slow
high wait on the system
when looker closer, the slapd process makes a very high volume of reads

This is fixed by a restart of the slapd (and rudder-jetty) services, but happens again later.

It seems to regularly trigger at 0:00, maybe because of the daily inventory update.

Seen at least on 6.2.10.

Actions

Copy link

Updated by Alexis Mousset about 3 years ago

Translation missing: en.field_tag_list set to Sponsored
Assignee set to Nicolas CHARLES
Severity set to Major - prevents use of part of Rudder | no simple workaround
User visibility set to Operational - other Techniques | Rudder settings | Plugins
Priority changed from 0 to 84

Actions

Copy link

Updated by Vincent MEMBRÉ about 3 years ago

Target version changed from 6.2.11 to 6.2.12
Priority changed from 84 to 83

Actions

Copy link

Updated by Nicolas CHARLES about 3 years ago

It's related to a lack of memory available for the system cache, so the ldap needs to look on drive rather than using cache
We could improve this situation by being less reliant on ldap for most common actions

Actions

Copy link