Bug #11896

All isotesters are broken due to guestfs freezing

Added by anonym 2016-11-04 12:15:11 . Updated 2016-11-09 10:12:06 .

Status:
Resolved
Priority:
High
Assignee:
Category:
Continuous Integration
Target version:
Start date:
2016-11-04
Due date:
% Done:

100%

Feature Branch:
Type of work:
Sysadmin
Blueprint:

Starter:
Affected tool:
Deliverable for:

Description

For all branches I see:

[...]
Feature: Untrusted partitions
  As a Tails user
  I don't want to touch other media than the one Tails runs from

  Scenario: Tails will not enable disk swap                       # features/untrusted_partitions.feature:6
    Given a computer                                              # features/step_definitions/common_steps.rb:122
    And I temporarily create a 100 MiB disk named "swap"          # features/step_definitions/common_steps.rb:139
23:14:21.566915970: libguestfs: trace: set_autosync true
23:14:21.567047844: libguestfs: trace: set_autosync = 0
23:14:21.567269186: libguestfs: trace: add_drive "/tmp/TailsToaster/TailsToasterStorage/swap" "format:qcow2"
23:14:21.567359204: libguestfs: trace: add_drive = 0
23:14:21.568114465: libguestfs: trace: launch
23:14:21.568203205: libguestfs: trace: get_tmpdir
23:14:21.568388347: libguestfs: trace: get_tmpdir = "/tmp/TailsToaster"
23:14:21.568504941: libguestfs: trace: get_backend_setting "force_tcg"
23:14:21.568615239: libguestfs: trace: get_backend_setting = NULL (error)
23:14:21.569047459: libguestfs: trace: get_cachedir
23:14:21.569195520: libguestfs: trace: get_cachedir = "/tmp/TailsToaster"
23:14:26.816791014: libguestfs: trace: get_backend_setting "gdb"
23:14:26.817021392: libguestfs: trace: get_backend_setting = NULL (error)


… and that’s when test suite run freezes and is aborted an hour later.

I actually saw this on a Debian Sid system while I was running the guestfs-based packages at version 1:1.32.7-1+b2, which was uploaded to jessie-backports just a few days ago, which would explain why it happened now. For me the issue seems to be fixed with 1.34.*, which unfortunately is not in jessie-packports or even stretch yet.


Subtasks


History

#1 Updated by anonym 2016-11-04 12:21:01

  • Description updated

#2 Updated by intrigeri 2016-11-04 14:16:36

This problem seems to have appeared when qemu got upgraded 1:2.7+dfsg-3~bpo8+1 to from 1:2.6+dfsg-3.1~bpo8+1 on Nov 3. Upgrading isotesters to jessie-backports’ 1:1.32.7-1~bpo8+1 does not fix it.

#3 Updated by intrigeri 2016-11-04 16:18:52

  • Status changed from Confirmed to In Progress
  • Assignee set to intrigeri

Backporting libguestfs 1.34 from sid. If that’s not enough I’ll temporarily upload the previous version of qemu (that did work for us) to our jenkins-slave APT suite.

#4 Updated by intrigeri 2016-11-04 17:48:29

  • % Done changed from 0 to 10

Reverted to Jessie’s libguestfs and qemu 1:2.6+dfsg-3.1~bpo8+1.

#5 Updated by intrigeri 2016-11-04 18:40:57

  • Assignee changed from intrigeri to anonym
  • % Done changed from 10 to 50
  • QA Check set to Ready for QA

Seems to be fixed. I might not be available to double-check.

#6 Updated by anonym 2016-11-09 10:12:06

  • Status changed from In Progress to Resolved
  • Assignee deleted (anonym)
  • % Done changed from 50 to 100
  • QA Check changed from Ready for QA to Pass

intrigeri wrote:
> Seems to be fixed. I might not be available to double-check.

Yup. Thanks!