Feature #11445

Give sysadmins tools to handle evergrowing reprepro database for time-based APT snapshots

Added by intrigeri 2016-05-18 12:05:41 . Updated 2016-06-01 12:20:10 .

Status:
Rejected
Priority:
Normal
Assignee:
Category:
Infrastructure
Target version:
Tails_2.5
Start date:
2016-05-18
Due date:
% Done:

100%

Feature Branch:
Type of work:
Code
Blueprint:

Starter:
Affected tool:
Deliverable for:
269

Description

Subtasks

History

#1 Updated by intrigeri 2016-05-18 12:42:37

  • Target version changed from Tails_2.4 to Tails_2.5

Problem well understood, hackish workarounds documented on the blueprint, icinga check in place, problem should not happen any time soon (the DB hasn’t actually grown since I compacted it a week ago, and it can grow from 1.5GB to 5.4GB and our thing will still work). FTR the DB grew to 5.4GB because we had daily snapshots (4 times a day) during 6+ months without any cleanup back then, but now we have some GC happening every day.

#2 Updated by intrigeri 2016-06-01 12:20:10

  • Status changed from In Progress to Rejected
  • Assignee deleted (intrigeri)
  • % Done changed from 0 to 100

Three weeks after db5.3_dump + db5.3_load, references.db has not grown at all, so in practice there’s no problem to be solved here.