Project

General

Profile

Actions
Copy link

action #155764

open

Consider switching to safer filesystems than ext2 in osd+o3

Added by okurz 9 months ago.

Status:
New
Priority:
Low
Assignee:
-
Category:
Feature requests
Target version:
QA - future
Start date:
Due date:
% Done:

0%

Estimated time:
Tags:
sqlite, cacheservice

Description

Motivation

In cases like #155716 we encountered problems due to the unsafe ext2 filesystem. Given that both filesystems as well as hardware have improved in performance I consider it reasonable to reconsider the original decision as of now and switch to a safer filesystem, possibly ext4 with journal kept on, for /var/lib/openqa. This can be combined with not recreating the filesystem on every boot to save bootup time.

Related issues 2 (0 open — 2 closed)

Related to openQA Infrastructure - action #162485: [alert] failed systemd service: openqa-worker-cacheservice on worker40.oqa.prg2.suse.org "Database has been corrupted: DBD::SQLite::db commit failed: disk I/O error" size:SResolvedokurz2024-06-19

Actions
Copied from openQA Project - action #155716: [alert] openqa-worker-cacheservice fails to start on worker29.oqa.prg2.suse.org with "Database has been corrupted: DBD::SQLite::db commit failed: disk I/O error" size:SResolvedmkittler2024-02-212024-03-07

Actions
Actions
Copy link
 #1

Updated by okurz 9 months ago

  • Copied from action #155716: [alert] openqa-worker-cacheservice fails to start on worker29.oqa.prg2.suse.org with "Database has been corrupted: DBD::SQLite::db commit failed: disk I/O error" size:S added
Actions
Copy link
 #2

Updated by okurz 5 months ago

  • Related to action #162485: [alert] failed systemd service: openqa-worker-cacheservice on worker40.oqa.prg2.suse.org "Database has been corrupted: DBD::SQLite::db commit failed: disk I/O error" size:S added
Actions
Copy link

Also available in: Atom PDF