[Bug 1089480] New: [Build 20180407] openQA btrfs_qgroups test reports bsc#1019614 - openSUSE Bugs - openSUSE Mailing Lists

newer
[Bug 1140564] New: kernel...

[Bug 1089480] New: [Build 20180407] openQA btrfs_qgroups test reports bsc#1019614

older
[Bug 1155510] New: iscsid service...

bugzilla_noreply＠novell.com

13 Apr 2018 13 Apr '18

09:05

http://bugzilla.opensuse.org/show_bug.cgi?id=1089480 Bug ID: 1089480 Summary: [Build 20180407] openQA btrfs_qgroups test reports bsc#1019614 Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: Other URL: http://9.101.18.90/tests/1105/modules/btrfs_qgroups/st eps/64 OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Kernel Assignee: kernel-maintainers@forge.provo.novell.com Reporter: normand@linux.vnet.ibm.com QA Contact: qa-bugs@suse.de Found By: --- Blocker: --- Created attachment 767068 --> http://bugzilla.opensuse.org/attachment.cgi?id=767068&action=edit disk_quota_exceeded.png openQA btrfs_qgroups test reports bsc#1019614 when tried on ppc64le. The reference to bsc#1019614 (I have no access to it) is related to https://github.com/os-autoinst/os-autoinst-distri-opensuse/blob/master/tests... The std output is full of: "dd: error writing 'e/file': Disk quota exceeded" as showned by attached png I have no link to openQA test on o3, because I made my test in an IBM openQA instance. (the extra_tests_filesystem is not enabled on o3 for ppc64le) I started to test with 20180402 snapshot, but still have it with 20180405 & 20180407. -- You are receiving this mail because: You are on the CC list for the bug.

Attachments:

attachment.htm (text/html — 3.4 KB)

Reply

Sign in to reply online Use email software

Show replies by date

bugzilla_noreply＠novell.com

13 Apr 13 Apr

09:06

New subject: [Bug 1089480] [Build 20180407] openQA btrfs_qgroups test reports bsc#1019614

http://bugzilla.opensuse.org/show_bug.cgi?id=1089480 Michel Normand changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |okurz@suse.com -- You are receiving this mail because: You are on the CC list for the bug.

Reply

Sign in to reply online Use email software

bugzilla_noreply＠novell.com

09:06

New subject: [Bug 1089480] [Build 20180407] openQA btrfs_qgroups test reports bsc#1019614

http://bugzilla.opensuse.org/show_bug.cgi?id=1089480 Michel Normand changed: What |Removed |Added ---------------------------------------------------------------------------- URL|http://9.101.18.90/tests/11 | |05/modules/btrfs_qgroups/st | |eps/64 | -- You are receiving this mail because: You are on the CC list for the bug.

Reply

Sign in to reply online Use email software

bugzilla_noreply＠novell.com

19 Jun 19 Jun

12:31

New subject: [Bug 1089480] [Build 20180407] openQA btrfs_qgroups test reports bsc#1019614

http://bugzilla.opensuse.org/show_bug.cgi?id=1089480 http://bugzilla.opensuse.org/show_bug.cgi?id=1089480#c3 Michel Normand changed: What |Removed |Added ---------------------------------------------------------------------------- Assignee|wqu@suse.com |normand@linux.vnet.ibm.com Flags|needinfo?(normand@linux.vne | |t.ibm.com) | --- Comment #3 from Michel Normand --- Created attachment 774481 --> http://bugzilla.opensuse.org/attachment.cgi?id=774481&action=edit btrfs_qgroup_show_before_and_after_rsync.png The attached image btrfs_qgroup_show_before_and_after_rsync.png is the results of suggested commands (1) I do not see differences in the output of the two calls of btrfs qgroup show. Do I need to try something else ? === $git diff tests/console/btrfs_qgroups.pm diff --git a/tests/console/btrfs_qgroups.pm b/tests/console/btrfs_qgroups.pm index 4a26771a..365e79d2 100644 --- a/tests/console/btrfs_qgroups.pm +++ b/tests/console/btrfs_qgroups.pm @@ -77,9 +77,12 @@ sub run { my $write_chunk = 'dd if=/dev/zero bs=1M count=40 of=e/file'; assert_script_run "for c in {1..2}; do $write_chunk; done", fail_message => 'bsc#1019614 overwriting same file should not exceed quota'; # write some more times to the same file to be sure - if (script_run "for c in {1..38}; do $write_chunk; done") { + if (script_run "for c in {1..38}; do $write_chunk || break; done") { record_soft_failure 'bsc#1019614'; } + assert_script_run "btrfs qgroup show -prce e"; + assert_script_run "sync"; + assert_script_run "btrfs qgroup show -prce e"; assert_script_run 'sync'; assert_script_run 'rm e/file', fail_message => 'bsc#993841'; # test exceeding real quota === -- You are receiving this mail because: You are on the CC list for the bug.

Reply

Sign in to reply online Use email software

bugzilla_noreply＠novell.com

23:26

New subject: [Bug 1089480] [Build 20180407] openQA btrfs_qgroups test reports bsc#1019614

http://bugzilla.opensuse.org/show_bug.cgi?id=1089480 http://bugzilla.opensuse.org/show_bug.cgi?id=1089480#c4 Wenruo Qu changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |wqu@suse.com Flags| |needinfo?(normand@linux.vne | |t.ibm.com) --- Comment #4 from Wenruo Qu --- So it's not the cache pages causing the problem. It's highly possible for the meta/data reserved space causing the problem. Since I'm no familiar with 64K sectorsize, I can't be 100% sure. Would you please removing all files of the testing subvolumes, and sync the fs using 'btrfs sub sync'. And finally try to fill the subvolume again, to see if we can still only reach that number. If so, we need extra info on the 64K sector size btrfs. And if possible, is there anyway to config the kernel on ppc64 to use 4K page size and re-test? Thanks, Qu -- You are receiving this mail because: You are on the CC list for the bug.

Reply

Sign in to reply online Use email software

bugzilla_noreply＠novell.com

31 Oct 31 Oct

15:28

New subject: [Bug 1089480] [Build 20180407] openQA btrfs_qgroups test reports bsc#1019614

http://bugzilla.opensuse.org/show_bug.cgi?id=1089480 http://bugzilla.opensuse.org/show_bug.cgi?id=1089480#c9 Michel Normand changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(normand@linux.vne | |t.ibm.com) | --- Comment #9 from Michel Normand --- I did not made any changes in related openQA test script or delivered kernel used by openQA test as requested in comment#4 The problem is still alive and ultimately make the test to randomly fail as per two examples for TW: snapshot 20191021: https://openqa.opensuse.org/tests/1063076#step/btrfs_qgroups/76 snapshot 20191030: https://openqa.opensuse.org/tests/1071384#step/btrfs_qgroups/76 https://openqa.opensuse.org/tests/1071384#downloads <= many captured data -- You are receiving this mail because: You are on the CC list for the bug.

Reply

Sign in to reply online Use email software

bugzilla_noreply＠novell.com

3 Nov 3 Nov

09:32

New subject: [Bug 1089480] [Build 20180407] openQA btrfs_qgroups test reports bsc#1019614

http://bugzilla.opensuse.org/show_bug.cgi?id=1089480 http://bugzilla.opensuse.org/show_bug.cgi?id=1089480#c10 --- Comment #10 from Wenruo Qu --- Btrfs qgroup limit calculation, is based on the principle of, never exceed the limit even it could return early EDQUOT. The main challenge for btrfs qgroup is, its reserved metadata space calculation is pretty unreliable. We over-reserve metadata space even for non-qgroup use cases. So current btrfs uses an optimistic method, if we're low on available space, we try to commit transaction in another thread (commit_kthread) to free some space. This behavior is not ensured to free enough space before we hit the limit, so we could have unexpected EDQUOT related to various factors like system load. The current behavior almost works fine for x86_64, but since we don't have ppc64 system to test, I guess it's more obvious to fail such optimistic tries. Would it be possible to mark that test as soft fail? As we don't have a good idea to completely fix it soon. -- You are receiving this mail because: You are on the CC list for the bug.

Reply

Sign in to reply online Use email software

bugzilla_noreply＠novell.com

18:16

New subject: [Bug 1089480] [Build 20180407] openQA btrfs_qgroups test reports bsc#1019614

http://bugzilla.opensuse.org/show_bug.cgi?id=1089480 http://bugzilla.opensuse.org/show_bug.cgi?id=1089480#c11 --- Comment #11 from Oliver Kurz --- (In reply to Wenruo Qu from comment #10)

Btrfs qgroup limit calculation, is based on the principle of, never exceed the limit even it could return early EDQUOT.

The main challenge for btrfs qgroup is, its reserved metadata space calculation is pretty unreliable. We over-reserve metadata space even for non-qgroup use cases.

So current btrfs uses an optimistic method, if we're low on available space, we try to commit transaction in another thread (commit_kthread) to free some space. This behavior is not ensured to free enough space before we hit the limit, so we could have unexpected EDQUOT related to various factors like system load.

The current behavior almost works fine for x86_64, but since we don't have ppc64 system to test, I guess it's more obvious to fail such optimistic tries.

Would it be possible to mark that test as soft fail? As we don't have a good idea to completely fix it soon.

The interesting thing is that the failing line is https://github.com/os-autoinst/os-autoinst-distri-opensuse/blob/master/tests... which is already within a "soft-fail workaround" section referencing bug 1113042 introduced in https://github.com/os-autoinst/os-autoinst-distri-opensuse/commit/443a07e954... by [Joaquín Rivera](https://github.com/jknphy) but certainly that can still be changed to record another workaround -- You are receiving this mail because: You are on the CC list for the bug.

Reply

Sign in to reply online Use email software

1642

Age (days ago)

2211

Last active (days ago)

Download

7 comments

1 participants

tags

participants (1)

bugzilla_noreply＠novell.com