MySQL Bugs: #37853: Test "funcs_1.processlist_val

Bug #37853	Test "funcs_1.processlist_val_ps" fails in various ways
Submitted:	3 Jul 2008 20:30	Modified:	6 Mar 2010 23:24
Reporter:	Joerg Bruehe	Email Updates:
Status:	Closed	Impact on me:	None
Category:	Tests: Server	Severity:	S3 (Non-critical)
Version:	5.1.26-rc	OS:	Other (Windows + others)
Assigned to:	Matthias Leich	CPU Architecture:	Any
Tags:	funcs_1, Tests, windows

Description:
This is in version 5.1.26-rc,
so the fix for bug#36033 is contained in the sources.

There are two different ways of failure:

1) "SELECT * FROM INFORMATION_SCHEMA.PROCESSLIST" sometimes reports non-zero time:
@@ -24,7 +24,7 @@
 USE test;
 SELECT * FROM INFORMATION_SCHEMA.PROCESSLIST;
 ID     USER    HOST    DB      COMMAND TIME    STATE   INFO
-<ID>   root    localhost       test    Execute 0       executing       SELECT * FROM INFORMATION_SCHEMA.PROCE
SSLIST
+<ID>   root    localhost       test    Execute 1       executing       SELECT * FROM INFORMATION_SCHEMA.PROCE
SSLIST
 SHOW FULL PROCESSLIST;
 Id     User    Host    db      Command Time    State   Info
 <ID>   root    localhost       test    Query   0       NULL    SHOW FULL PROCESSLIST

In 5.1.26-rc, till now I had this on BSD 6.0 (x86_64) and i5os (32 bit only).

2) The replacement of the host name by a constant is incomplete:
@@ -24,10 +24,10 @@
 USE test;
 SELECT * FROM INFORMATION_SCHEMA.PROCESSLIST;
 ID     USER    HOST    DB      COMMAND TIME    STATE   INFO
-<ID>   root    localhost       test    Execute 0       executing       SELECT * FROM INFORMATION_SCHEMA.PROCE
SSLIST
+<ID>   root    localhost:25591 test    Execute 0       executing       SELECT * FROM INFORMATION_SCHEMA.PROCE
SSLIST
 SHOW FULL PROCESSLIST;
 Id     User    Host    db      Command Time    State   Info
-<ID>   root    localhost       test    Query   0       NULL    SHOW FULL PROCESSLIST
+<ID>   root    localhost:25591 test    Query   0       NULL    SHOW FULL PROCESSLIST
 USE information_schema;
 SELECT * FROM INFORMATION_SCHEMA.PROCESSLIST;
 ID     USER    HOST    DB      COMMAND TIME    STATE   INFO

Similar in lines 110/111, 114/115, and 118/119 of the result file.
For all these commands, I do not see a "--replace_column" for the host name.
Happens in all runs on Windows (with varying port numbers).

How to repeat:
Run the "funcs_1" suite, especially on Windows.

Suggested fix:
For 1), check whether this is acceptable;
if yes, add "--replace_column".

For 2), add "--replace_column".

I forgot the third way of failure:

The command (start) may be reported different, like here
(incomplete, because of that 21 kB command):

  SHOW PROCESSLIST;
  Id    User    Host    db      Command Time    State   Info
  <ID>  root    localhost       information_schema      <COMMAND>       <TIME>  <STATE> SHOW PROCESSLIST
! <ID>  ddicttestuser1  localhost       information_schema      <COMMAND>       <TIME>  <STATE> SELECT sleep(2),'BEGIN this is the representative of a very long statement.this is the representativ

  ----- switch to connection ddicttestuser1 (user = ddicttestuser1) -----
  # Pull("reap") the result set from the monster statement executed with "send".
--- 107,122
  # Sleep some time
  SELECT * FROM INFORMATION_SCHEMA.PROCESSLIST;
  ID    USER    HOST    DB      COMMAND TIME    STATE   INFO
! <ID>  ddicttestuser1  localhost       information_schema      <COMMAND>       <TIME>  <STATE> SELECT sleep(2.5),'Command time'
  <ID>  root    localhost       information_schema      <COMMAND>       <TIME>  <STATE> SELECT * FROM INFORMATION_SCHEMA.PROCESSLIST
  SHOW FULL PROCESSLIST;
  Id    User    Host    db      Command Time    State   Info
  <ID>  root    localhost       information_schema      <COMMAND>       <TIME>  <STATE> SHOW FULL PROCESSLIST
! <ID>  ddicttestuser1  localhost       information_schema      <COMMAND>       <TIME>  <STATE> SELECT sleep(2.5),'Command time'
  SHOW PROCESSLIST;
  Id    User    Host    db      Command Time    State   Info
  <ID>  root    localhost       information_schema      <COMMAND>       <TIME>  <STATE> SHOW PROCESSLIST
! <ID>  ddicttestuser1  localhost       information_schema      <COMMAND>       <TIME>  <STATE> SELECT sleep(2.5),'Command time'

  ----- switch to connection ddicttestuser1 (user = ddicttestuser1) -----
  # Pull("reap") the result set from the monster statement executed with "send".

This happened on HP-UX 11.23 (IA64).

I have not analyzed the test to check whether this is a timing issue, or a functional problem.

This bug (mostly the differences caused by the port)
is unexpected, because the fix for
Bug#36033 Test funcs_1.processlist_val_ps fails on Windows
had at least to avoid the differences around port.
Maybe a unfortunate merge removed the fix.

Hi Triage team,
please give this bug sufficient priority.
Thank you.
1. The printing of the port causes that all runs
   on Windows fail.
   - This affects the tests run by the build team.
   - There is no impact on usual Pushbuild because
     all funcs_1 runs happen on Linux boxes.
2. The assumptions about timing seem to be too optimistic
   and affect all platforms.
   = Will cause random differences in Pushbuild
     and build team tests on every platform
     whenever the load on the testing box is high.

A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/49295

A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/49595

Ok to push.

A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/51224

A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/51346

A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/51563

A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/51568

A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/51569

A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/51575

A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

  http://lists.mysql.com/commits/51579

2763 Marc Alff	2008-08-13 [merge]
      Merge 6.0-bugteam -> local bugfix branch

Pushed to
5.1-bugteam
6.0-bugteam

mysql-5.1 $ bzr log -r-300.. -m 37853
    ------------------------------------------------------------
    revno: 2646.5.42
    revision-id: mleich@mysql.com-20080813194454-e2mwdis01c309yca
    parent: mleich@mysql.com-20080813194221-u93w7gf6pbb52bc8
    committer: Matthias Leich <mleich@mysql.com>
    branch nick: mysql-5.1-bugteam-push
    timestamp: Wed 2008-08-13 21:44:54 +0200
    message:
      Fix for Bug#37853
          Test "funcs_1.processlist_val_ps" fails in various ways
      + corrections of logic in poll routines
      + minor improvements

Test case changes. No changelog entry needed.

Pushed into 6.0.8-alpha  (revid:mleich@mysql.com-20080813194454-e2mwdis01c309yca) (version source revid:v.narayanan@sun.com-20080908113103-2e4vdpm695vk3ji4) (pib:3)

Pushed into 6.0.14-alpha (revid:alik@sun.com-20100216101445-2ofzkh48aq2e0e8o) (version source revid:jon.hauglid@sun.com-20100114131510-hqdndvv6mrws762w) (merge vers: 6.0.14-alpha) (pib:16)

Pushed into mysql-next-mr (revid:alik@sun.com-20100216101208-33qkfwdr0tep3pf2) (version source revid:jon.hauglid@sun.com-20100114130324-qtvz8fshxso9yzz6) (pib:16)

Pushed into 5.5.3-m3 (revid:alik@sun.com-20100306103849-hha31z2enhh7jwt3) (version source revid:vvaintroub@mysql.com-20100216221947-luyhph0txl2c5tc8) (merge vers: 5.5.99-m3) (pib:16)

No changelog entry needed.