Problems setting patchbots

114 views
Skip to first unread message

mmarco

unread,
Dec 6, 2019, 7:04:06 AM12/6/19
to sage-devel
I am setting up several virtual machines with different distros to run patchbots on them. In particular I want to use different distros (including less usual ones) to get more diversity in the patchbot set.

However, I am finding some problems:

A debian machine is returning the following error when testing the base install:

sage -t --long --warn-long 41.0 src/sage/tests/gap_packages.py
**********************************************************************
File "src/sage/tests/gap_packages.py", line 8, in sage.tests.gap_packages
Failed example:
    test_packages(pkgs, only_failures=True)    # optional - gap_packages
Expected:
      Status   Package   GAP Output
    +--------+---------+------------+
Got:
    #I  method installed for Matrix matches more than one declaration
      Status   Package   GAP Output
    +--------+---------+------------+
**********************************************************************
1 item had failures:
   1 of   5 in sage.tests.gap_packages
    [11 tests, 1 failure, 3.59 s]
----------------------------------------------------------------------
sage -t --long --warn-long 41.0 src/sage/tests/gap_packages.py  # 1 doctest failed
----------------------------------------------------------------------
Total time for all tests: 4752.5 seconds
    cpu time: 7926.2 seconds
    cumulative wall time: 9116.0 seconds


In particular, it is this machine:



Another one, running OpenSUSE, is also givin the same error, but besides, doesn't seem to be sending reports to the server:





Any clues?

Dima Pasechnik

unread,
Dec 6, 2019, 7:07:34 AM12/6/19
to sage-devel
On Fri, Dec 6, 2019 at 12:04 PM mmarco <mma...@unizar.es> wrote:
>
> I am setting up several virtual machines with different distros to run patchbots on them. In particular I want to use different distros (including less usual ones) to get more diversity in the patchbot set.
>
> However, I am finding some problems:
>
> A debian machine is returning the following error when testing the base install:
>
> sage -t --long --warn-long 41.0 src/sage/tests/gap_packages.py
> **********************************************************************
> File "src/sage/tests/gap_packages.py", line 8, in sage.tests.gap_packages
> Failed example:
> test_packages(pkgs, only_failures=True) # optional - gap_packages
> Expected:
> Status Package GAP Output
> +--------+---------+------------+
> Got:
> #I method installed for Matrix matches more than one declaration
> Status Package GAP Output
> +--------+---------+------------+
> **********************************************************************
> 1 item had failures:
> 1 of 5 in sage.tests.gap_packages
> [11 tests, 1 failure, 3.59 s]
> ----------------------------------------------------------------------
> sage -t --long --warn-long 41.0 src/sage/tests/gap_packages.py # 1 doctest failed


this is a known (minor) error, to be fixed in upcoming GAP 4.11
It only pops up if gap_packages optional packages is installed.
One can also update the relevant parts of gap_packages to fix it,
as I already mentioned here (or on sage-release)

> ----------------------------------------------------------------------
> Total time for all tests: 4752.5 seconds
> cpu time: 7926.2 seconds
> cumulative wall time: 9116.0 seconds
>
>
> In particular, it is this machine:
>
> https://patchbot.sagemath.org/ticket/?machine=debian&machine=10.2&machine=x86_64&machine=4.19.0-6-amd64&machine=localhost&status=open
>
>
> Another one, running OpenSUSE, is also givin the same error, but besides, doesn't seem to be sending reports to the server:
>
> https://patchbot.sagemath.org/log/0/Linux/#1%20SMP%20Wed%20Nov%2013%2007:50:15%20UTC%202019%20(6e1aaad)/x86_64/4.12.14-lp151.28.32-default/pb-unizar-opensuse/2019-12-06%2011:48:21
>
>
>
>
> Any clues?
>
> --
> You received this message because you are subscribed to the Google Groups "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/91ff2bed-d347-4369-b292-e660ea62adea%40googlegroups.com.

Frédéric Chapoton

unread,
Dec 6, 2019, 9:54:10 AM12/6/19
to sage-devel
The first problem is a known one, you can uninstall gap_packages to fix it.

The second may be a bug in the patchbot server itself. I have no time to investigate. Maybe the # in the patchbot name is not a good idea (random guess) ?

F

mmarco

unread,
Dec 8, 2019, 7:27:56 AM12/8/19
to sage-devel

The second may be a bug in the patchbot server itself. I have no time to investigate. Maybe the # in the patchbot name is not a good idea (random guess) ?

F

I see. Is there a command to uninstall a package? or should I edit some file manually?

I will try avoiding the number in the name.

mmarco

unread,
Dec 10, 2019, 12:47:08 PM12/10/19
to sage-devel
For completeness, the problem of the log not being sent is also present in another VM running arch:



Maybe some trouble of the patchbot recognizing the distribution name?

mmarco

unread,
Dec 17, 2019, 10:55:27 AM12/17/19
to sage-devel

The second may be a bug in the patchbot server itself. I have no time to investigate. Maybe the # in the patchbot name is not a good idea (random guess) ?


It looks like that could be the culprit, since It also happens in other machines. Any idea on how it could be solved? 

Frédéric Chapoton

unread,
Dec 17, 2019, 10:57:48 AM12/17/19
to sage-devel
Apparenlty, the # symbol was (at least part of) the problem. I pushed a fix a few minutes ago, not yet tested. You can git pull the latest patchbot sources.

F
Reply all
Reply to author
Forward
0 new messages