Merge ~athos-ribeiro/ubuntu/+source/pglogical:increase-timeout into ubuntu/+source/pglogical:ubuntu/devel

[Athos Ribeiro]

Increase timeout on flaky tests.

PPA: https://launchpad.net/~athos-ribeiro/+archive/ubuntu/pglogical-timeouts/+packages

Waiting for the PPA's dep8 test results.

[Sergio Durigan Junior (sergiodj)]

Thanks for the MP, Athos.

Were you able to reproduce the failure locally? I spent a few hours last Friday trying to do so, but failed. At the time I thought the discrepancy could be caused by the fact that I'd rebuilt the package locally before running the tests, so I did a no-change upload in the hopes that it could make a difference, but unfortunately the test is still failing... I also built the package in a PPA, but it's hard to trigger the dep8 test using postgresql-15 as one of the triggers.

Anyway, I'm just curious to see if you were able to reproduce the failure and confirm that it's indeed a flaky test. Thanks!

review: Needs Information

[Athos Ribeiro (athos-ribeiro)]

Thanks, Sergio! All tests are passing with the new timeout:

https://autopkgtest.ubuntu.com/results/autopkgtest-lunar-athos-ribeiro-pglogical-timeouts/lunar/ppc64el/p/pglogical/20221213_201748_7c463@/log.gz

https://autopkgtest.ubuntu.com/results/autopkgtest-lunar-athos-ribeiro-pglogical-timeouts/lunar/armhf/p/pglogical/20221213_192348_d9e5a@/log.gz

https://autopkgtest.ubuntu.com/results/autopkgtest-lunar-athos-ribeiro-pglogical-timeouts/lunar/arm64/p/pglogical/20221213_201851_10806@/log.gz

https://autopkgtest.ubuntu.com/results/autopkgtest-lunar-athos-ribeiro-pglogical-timeouts/lunar/amd64/p/pglogical/20221213_202119_7a5cc@/log.gz

I could not reproduce it locally, but to ensure my PPA pg-15 build is not the variant fixing the test runs, I will revert the pglogical changes for the timeout values and verify they indeed fail (this, combined with the d/changelog entries and d/patches changing timeout values and disabling flaky tests should be enough to confirm this is a resources issue).

[Sergio Durigan Junior (sergiodj)]

On Tuesday, December 13 2022, Athos Ribeiro wrote:

> Thanks, Sergio! All tests are passing with the new timeout:
>
> https://autopkgtest.ubuntu.com/results/autopkgtest-lunar-athos-ribeiro-pglogical-timeouts/lunar/ppc64el/p/pglogical/20221213_201748_7c463@/log.gz
>
> https://autopkgtest.ubuntu.com/results/autopkgtest-lunar-athos-ribeiro-pglogical-timeouts/lunar/armhf/p/pglogical/20221213_192348_d9e5a@/log.gz
>
> https://autopkgtest.ubuntu.com/results/autopkgtest-lunar-athos-ribeiro-pglogical-timeouts/lunar/arm64/p/pglogical/20221213_201851_10806@/log.gz
>
> https://autopkgtest.ubuntu.com/results/autopkgtest-lunar-athos-ribeiro-pglogical-timeouts/lunar/amd64/p/pglogical/20221213_202119_7a5cc@/log.gz
>
> I could not reproduce it locally, but to ensure my PPA pg-15 build is
> not the variant fixing the test runs, I will revert the pglogical
> changes for the timeout values and verify they indeed fail (this,
> combined with the d/changelog entries and d/patches changing timeout
> values and disabling flaky tests should be enough to confirm this is a
> resources issue).

Thanks, Athos. That's great news :-). Let's wait and see what happens
with the reverted changes, then. But if you feel like going ahead and
uploading this to the archive directly, I wouldn't oppose.

Cheers,

--
Sergio
GPG key ID: E92F D0B3 6B14 F1F4 D8E0 EB2F 106D A1C8 C3CB BF14

[Athos Ribeiro (athos-ribeiro)]

armhf passed, as the actual results we got in the archive: https://autopkgtest.ubuntu.com/results/autopkgtest-lunar-athos-ribeiro-pglogical-timeouts/lunar/armhf/p/pglogical/20221213_224806_7b1c5@/log.gz

amd64 and ppc64el failed, just as the actual results in the archive (same failure): https://autopkgtest.ubuntu.com/results/autopkgtest-lunar-athos-ribeiro-pglogical-timeouts/lunar/amd64/p/pglogical/20221213_233338_bf70c@/log.gz and https://autopkgtest.ubuntu.com/results/autopkgtest-lunar-athos-ribeiro-pglogical-timeouts/lunar/ppc64el/p/pglogical/20221213_234115_78ca0@/log.gz

differently from the results for the -proposed packages (and the tests above), arm64 passed after reverting the changes: https://autopkgtest.ubuntu.com/results/autopkgtest-lunar-athos-ribeiro-pglogical-timeouts/lunar/arm64/p/pglogical/20221214_002242_acdbf@/log.gz.

Still, considering the failures for ppc64el and amd64, these do seem to be flaky tests. IMO, we should merge and upload this. If the actual results are still flaky, we could proceed to disable this specific test.

[Sergio Durigan Junior (sergiodj)]

I guess that settles it, then. Thanks for the thorough investigation. SHIP IT!

review: Approve

[git-ubuntu bot (git-ubuntu-bot)]

Approvers: athos-ribeiro, sergiodj
Uploaders: athos-ribeiro, sergiodj
MP auto-approved

review: Approve

[Athos Ribeiro (athos-ribeiro)]

Thanks!

Uploaded

Uploading to ubuntu (via ftp to upload.ubuntu.com):
  Uploading pglogical_2.4.2-2ubuntu1.dsc: done.
  Uploading pglogical_2.4.2-2ubuntu1.debian.tar.xz: done.
  Uploading pglogical_2.4.2-2ubuntu1_source.buildinfo: done.
  Uploading pglogical_2.4.2-2ubuntu1_source.changes: done.
Successfully uploaded packages.

diff --git a/debian/changelog b/debian/changelog
index ba6bf94..07be017 100644
--- a/debian/changelog
+++ b/debian/changelog
@@ -1,3 +1,10 @@
+pglogical (2.4.2-2ubuntu1) lunar; urgency=medium
+
+  * d/p/test_increase_timeouts.patch: increase timeout for flaky
+    row_filter test.
+
+ -- Athos Ribeiro <[e-mail]>  Tue, 13 Dec 2022 10:32:25 -0300
+
 pglogical (2.4.2-2build1) lunar; urgency=medium
 
   * No-change rebuild to attempt migration.
diff --git a/debian/control b/debian/control
index a52a7b7..705920f 100644
--- a/debian/control
+++ b/debian/control
@@ -1,7 +1,8 @@
 Source: pglogical
 Section: database
 Priority: optional
-Maintainer: Debian PostgreSQL Maintainers <[e-mail]>
+Maintainer: Ubuntu Developers <[e-mail]>
+XSBC-Original-Maintainer: Debian PostgreSQL Maintainers <[e-mail]>
 Uploaders: Michael Banck <[e-mail]>, Marco Nenciarini <[e-mail]>
 Build-Depends:
  debhelper-compat (= 13),
diff --git a/debian/control.in b/debian/control.in
index b2eb307..7366501 100644
--- a/debian/control.in
+++ b/debian/control.in
@@ -1,7 +1,8 @@
 Source: pglogical
 Section: database
 Priority: optional
-Maintainer: Debian PostgreSQL Maintainers <[e-mail]>
+Maintainer: Ubuntu Developers <[e-mail]>
+XSBC-Original-Maintainer: Debian PostgreSQL Maintainers <[e-mail]>
 Uploaders: Michael Banck <[e-mail]>, Marco Nenciarini <[e-mail]>
 Build-Depends:
  debhelper-compat (= 13),
diff --git a/debian/patches/test_increase_timeouts.patch b/debian/patches/test_increase_timeouts.patch
index 55cb005..e596873 100644
--- a/debian/patches/test_increase_timeouts.patch
+++ b/debian/patches/test_increase_timeouts.patch
@@ -59,7 +59,7 @@ Index: pglogical/sql/row_filter.sql
  -- wait for the initial data to copy
  BEGIN;
 -SET LOCAL statement_timeout = '10s';
-+SET LOCAL statement_timeout = '60s';
++SET LOCAL statement_timeout = '120s';
  SELECT pglogical.wait_for_subscription_sync_complete('test_subscription');
  COMMIT;
  
@@ -72,7 +72,7 @@ Index: pglogical/expected/row_filter.out
  -- wait for the initial data to copy
  BEGIN;
 -SET LOCAL statement_timeout = '10s';
-+SET LOCAL statement_timeout = '60s';
++SET LOCAL statement_timeout = '120s';
  SELECT pglogical.wait_for_subscription_sync_complete('test_subscription');
   wait_for_subscription_sync_complete 
  -------------------------------------
@@ -85,7 +85,7 @@ Index: pglogical/expected/row_filter_1.out
  -- wait for the initial data to copy
  BEGIN;
 -SET LOCAL statement_timeout = '10s';
-+SET LOCAL statement_timeout = '60s';
++SET LOCAL statement_timeout = '120s';
  SELECT pglogical.wait_for_subscription_sync_complete('test_subscription');
   wait_for_subscription_sync_complete 
  -------------------------------------