Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.debian.bugs.dist > #1214019
| From | Craig Small <csmall@dropbear.xyz> |
|---|---|
| Newsgroups | linux.debian.bugs.dist |
| Subject | Bug#1072533: procps: flaky autopkgtest: |
| Date | 2024-09-23 10:30 +0200 |
| Message-ID | <JpMCd-efdD-5@gated-at.bofh.it> (permalink) |
| References | (1 earlier) <ILm0y-nAE-5@gated-at.bofh.it> <J20mZ-3fyM-5@gated-at.bofh.it> <J2cHv-3qgF-1@gated-at.bofh.it> <ILm0y-nAE-5@gated-at.bofh.it> <J2cHv-3qgF-1@gated-at.bofh.it> |
| Organization | linux.* mail to news gateway |
[Multipart message — attachments visible in raw view] - view raw
On Sat, 20 Jul 2024 at 17:24, Paul Gevers <elbrus@debian.org> wrote:
> I can give you access to a testbed on our infrastructure where you can
> play in a production environment exactly like the tests run. We'd need
> to align the date/time, as in the current way that I can provide this
> service, I need to be on-line at the same time.
>
Hi Paul,
I basically need to see what vmstat is doing, there is something very odd
with the lxc environment
that CI uses and its making vmstat misbehave; I'm not even seeing it report
any errors except a return value.
The CI reports:
31s test_no_arguments
31s [1;31mASSERT: [0mvalue line
31s [1;31mshunit2:ERROR [0m test_no_arguments() returned non-zero return
code.
If the output is incorrect it looks like this:
Running
/home/csmall/Projects/procps/procps/testsuite/vmstat.test/vmstat.exp ...
FAIL: vmstat with no arguments
If I force vmstat to return non zero it looks like:
Running
/home/csmall/Projects/procps/procps/testsuite/vmstat.test/vmstat.exp ...
=== vmstat Summary ===
# of expected passes 4
# of unsupported tests 2
Which is odd looking but correct because the test script is looking for the
output.
Also the -a option doesn't fail, only the empty option, which narrows it
down further.
This means we're down to what is found in new_format() which 'vmstat' runs
but 'vmstat -a' does not.
Which is lines like:
V( 4) = unitConvert((a_option?MEMv(mem_INA):MEMv(mem_BUF)));
V( 5) = unitConvert((a_option?MEMv(mem_ACT):MEMv(mem_CAC)));
mem_BUF directly comes from /proc/meminfo line Buffers
mem_CAC is calculated by Cached + SReclaimable
Something is not right here, we've had issues in the past where lxc doesn't
quite put the right values in.
If possible, could you do the following?
$ grep -E '^(Buffers|Cached|SReclaimable):' /proc/meminfo ; vmstat
Buffers: 1221344 kB
Cached: 14260336 kB
SReclaimable: 885284 kB
procs -----------memory---------- ---swap-- -----io---- -system--
-------cpu-------
r b swpd free buff cache si so bi bo in cs us sy id
wa st gu
1 0 13416 7994060 1221344 15145620 0 0 29 107 1372 22 8 1
91 0 0 0
buff column is the same as Buffers line (1221344)
cache column is Cached + SReclaimable (14260336 + 885284 = 15145620)
I tried it in a lxc container and vmstat works, but (correctly) shows 0 for
Buffers and SReclaimable.
- Craig
Back to linux.debian.bugs.dist | Previous | Next — Next in thread | Find similar | Unroll thread
Bug#1072533: procps: flaky autopkgtest: Craig Small <csmall@dropbear.xyz> - 2024-09-23 10:30 +0200
Bug#1072533: procps: flaky autopkgtest: Paul Gevers <elbrus@debian.org> - 2024-09-23 21:20 +0200
Bug#1072533: procps: flaky autopkgtest: Craig Small <csmall@debian.org> - 2024-09-25 12:20 +0200
Bug#1072533: procps: flaky autopkgtest: Craig Small <csmall@debian.org> - 2024-09-29 10:10 +0200
csiph-web