Re: [sage-devel] Re: Graceful shutdown of patchbot

[Jeroen Demeyer]

On 2016-09-21 13:55, Frédéric Chapoton wrote:

This is because the patchbot called with "sage -patchbot" needs
currently to be launched from the branch of
https://trac.sagemath.org/ticket/20736


Well, *you* (or anybody else) could help by reviewing that ticket.

--
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+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Graceful shutdown of patchbot

[John Cremona]

On 21 September 2016 at 12:55, Frédéric Chapoton  wrote:
> This is because the patchbot called with "sage -patchbot" needs currently to
> be launched from the branch of https://trac.sagemath.org/ticket/20736
>
> So you need to switch back to this branch.. This is very inconvenient, but
> as long as this ticket is not closed, thiis is like that. I would strongly
> advise that you use the other way to launch the patchbot, as explained in
>
> https://wiki.sagemath.org/buildbot/details

OK, so that worked after I replaced the line found there with

./sage -python -m sage_patchbot.patchbot --sage-root=$PWD

(starting in the Sage root directory).  I suggest that the wiki is
changed to remove the way which just failed, and also to use "sage
-python" instead of "python".

So patchbot fermat is running again, currently testing the base (which
I thought it best not to skip in the circumstances).

John

>
> Frederic
>
> Le mercredi 21 septembre 2016 11:43:36 UTC+2, John Cremona a écrit :
>>
>> Related question:  My patchbot-running machine was rebooted
>> ungracefully.  It's on a git branch called ticket_merged, and ./sage
>> -patchbot fails with
>>
>> /home/jec/sagedev/src/bin/sage: line 271:
>> /home/jec/sagedev/local/bin/patchbot/patchbot.py: No such file or
>> directory
>>
>> How to recover?  I was running with everything on default settings.
>>
>> John
>>
>> On 15 September 2016 at 10:03, Kwankyu Lee  wrote:
>> >
>> >> There is an option --count to say how many ticket to tests before to
>> >> stop.
>> >> you may want to use that.
>> >> You may even change it in your json config file during the patchbot run
>> >> (not tested if this works to stop the bot)
>> >
>> >
>> > Ok. I will try that. Thanks.
>> >
>> >> But Ctrl-C should not be very problematic. Could you tell what kind of
>> >> problems you have seen ?
>> >
>> >
>> > After a shutdown by ctrl-c, a subsequent doctesting used an old base and
>> > commits:
>> >
>> > Commit: b4c6cd222f4789617f15fac8ddec6c7e8d6e9f99 (7.0.beta0 + 5 commits)
>> >
>> > while in the previous testing, it was correctly:
>> >
>> > Commit: c57069e9c66ad4f28038744f370fe6aac9c574d0 (7.4.beta4 + 9 commits)
>> >
>> > I cannot understand exactly how this happened...
>> >
>> >
>> >
>> >
>> >
>> > Frederic
>> >
>> > Le jeudi 15 septembre 2016 09:49:13 UTC+2, Kwankyu Lee a écrit :
>> > Dear all,
>> >
>> >
>> > A patchbot is supposed to run forever, but realistically I should stop
>> > it
>> > from time to time. Pushing ctrl-c stops the patchbot abruptly, and the
>> > subsequent run seems sometimes to show somewhat erroneous behavior,
>> > perhaps
>> > due to the spurious state of the files.
>> >
>> >
>> > Is there a way to stop the patchbot gracefully? Or am I worrying too
>> > much
>> > and ctrl-c is just ok?
>> >
>> >
>> > On Thursday, September 15, 2016 at 10:00:47 AM UTC+2, Frédéric Chapoton
>> > wrote:
>> >>
>> >> There is an option --count to say how many ticket to tests before to
>> >> stop.
>> >> you may want to use that.
>> >>
>> >> You may even change it in your json config file during the patchbot run
>> >> (not tested if this works to stop the bot)
>> >
>> >
>> > Ok. I will try that. Thanks.
>> >
>> >>
>> >> But Ctrl-C should not be very problematic. Could you tell what kind of
>> >> problems you have seen ?
>> >
>> >
>> > I cannot say exactly. After a shutdown by ctrl-c, a subsequent
>> > doctesting
>> > used an old base:
>> >
>> > Commit: b4c6cd222f4789617f15fac8ddec6c7e8d6e9f99 (7.0.beta0 + 5 commits)
>> >
>> > while
>> >
>> >>
>> >>
>> >> Frederic
>> >>
>> >> Le jeudi 15 septembre 2016 09:49:13 UTC+2, Kwankyu Lee a écrit :
>> >>>
>> >>> Dear all,
>> >>>
>> >>> A patchbot is supposed to run forever, but realistically I should stop
>> >>> it
>> >>> from time to time. Pushing ctrl-c stops the patchbot abruptly, and the
>> >>> subsequent run seems sometimes to show somewhat erroneous behavior,
>> >>> perhaps
>> >>> due to the spurious state of the files.
>> >>>
>> >>> Is there a way to stop the patchbot gracefully? Or am I worrying too
>> >>> much
>> >>> and ctrl-c is just ok?
>> >
>> > --
>> > 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 post to this group, send email to sage-...@googlegroups.com.
>> > Visit this group at https://groups.google.com/group/sage-devel.
>> > For more options, visit https://groups.google.com/d/optout.
>
> --
> 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+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit 

Re: [sage-devel] Re: Graceful shutdown of patchbot

[Frédéric Chapoton]

This is because the patchbot called with "sage -patchbot" needs currently 
to be launched from the branch of https://trac.sagemath.org/ticket/20736

So you need to switch back to this branch.. This is very inconvenient, but 
as long as this ticket is not closed, thiis is like that. I would strongly 
advise that you use the other way to launch the patchbot, as explained in

https://wiki.sagemath.org/buildbot/details

Frederic

Le mercredi 21 septembre 2016 11:43:36 UTC+2, John Cremona a écrit :
>
> Related question:  My patchbot-running machine was rebooted 
> ungracefully.  It's on a git branch called ticket_merged, and ./sage 
> -patchbot fails with 
>
> /home/jec/sagedev/src/bin/sage: line 271: 
> /home/jec/sagedev/local/bin/patchbot/patchbot.py: No such file or 
> directory 
>
> How to recover?  I was running with everything on default settings. 
>
> John 
>
> On 15 September 2016 at 10:03, Kwankyu Lee  > wrote: 
> > 
> >> There is an option --count to say how many ticket to tests before to 
> stop. 
> >> you may want to use that. 
> >> You may even change it in your json config file during the patchbot run 
> >> (not tested if this works to stop the bot) 
> > 
> > 
> > Ok. I will try that. Thanks. 
> > 
> >> But Ctrl-C should not be very problematic. Could you tell what kind of 
> >> problems you have seen ? 
> > 
> > 
> > After a shutdown by ctrl-c, a subsequent doctesting used an old base and 
> > commits: 
> > 
> > Commit: b4c6cd222f4789617f15fac8ddec6c7e8d6e9f99 (7.0.beta0 + 5 commits) 
> > 
> > while in the previous testing, it was correctly: 
> > 
> > Commit: c57069e9c66ad4f28038744f370fe6aac9c574d0 (7.4.beta4 + 9 commits) 
> > 
> > I cannot understand exactly how this happened... 
> > 
> > 
> > 
> > 
> > 
> > Frederic 
> > 
> > Le jeudi 15 septembre 2016 09:49:13 UTC+2, Kwankyu Lee a écrit : 
> > Dear all, 
> > 
> > 
> > A patchbot is supposed to run forever, but realistically I should stop 
> it 
> > from time to time. Pushing ctrl-c stops the patchbot abruptly, and the 
> > subsequent run seems sometimes to show somewhat erroneous behavior, 
> perhaps 
> > due to the spurious state of the files. 
> > 
> > 
> > Is there a way to stop the patchbot gracefully? Or am I worrying too 
> much 
> > and ctrl-c is just ok? 
> > 
> > 
> > On Thursday, September 15, 2016 at 10:00:47 AM UTC+2, Frédéric Chapoton 
> > wrote: 
> >> 
> >> There is an option --count to say how many ticket to tests before to 
> stop. 
> >> you may want to use that. 
> >> 
> >> You may even change it in your json config file during the patchbot run 
> >> (not tested if this works to stop the bot) 
> > 
> > 
> > Ok. I will try that. Thanks. 
> > 
> >> 
> >> But Ctrl-C should not be very problematic. Could you tell what kind of 
> >> problems you have seen ? 
> > 
> > 
> > I cannot say exactly. After a shutdown by ctrl-c, a subsequent 
> doctesting 
> > used an old base: 
> > 
> > Commit: b4c6cd222f4789617f15fac8ddec6c7e8d6e9f99 (7.0.beta0 + 5 commits) 
> > 
> > while 
> > 
> >> 
> >> 
> >> Frederic 
> >> 
> >> Le jeudi 15 septembre 2016 09:49:13 UTC+2, Kwankyu Lee a écrit : 
> >>> 
> >>> Dear all, 
> >>> 
> >>> A patchbot is supposed to run forever, but realistically I should stop 
> it 
> >>> from time to time. Pushing ctrl-c stops the patchbot abruptly, and the 
> >>> subsequent run seems sometimes to show somewhat erroneous behavior, 
> perhaps 
> >>> due to the spurious state of the files. 
> >>> 
> >>> Is there a way to stop the patchbot gracefully? Or am I worrying too 
> much 
> >>> and ctrl-c is just ok? 
> > 
> > -- 
> > 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 post to this group, send email to sage-...@googlegroups.com 
> . 
> > Visit this group at https://groups.google.com/group/sage-devel. 
> > For more options, visit https://groups.google.com/d/optout. 
>

-- 
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+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Graceful shutdown of patchbot

[John Cremona]

Related question:  My patchbot-running machine was rebooted
ungracefully.  It's on a git branch called ticket_merged, and ./sage
-patchbot fails with

/home/jec/sagedev/src/bin/sage: line 271:
/home/jec/sagedev/local/bin/patchbot/patchbot.py: No such file or
directory

How to recover?  I was running with everything on default settings.

John

On 15 September 2016 at 10:03, Kwankyu Lee  wrote:
>
>> There is an option --count to say how many ticket to tests before to stop.
>> you may want to use that.
>> You may even change it in your json config file during the patchbot run
>> (not tested if this works to stop the bot)
>
>
> Ok. I will try that. Thanks.
>
>> But Ctrl-C should not be very problematic. Could you tell what kind of
>> problems you have seen ?
>
>
> After a shutdown by ctrl-c, a subsequent doctesting used an old base and
> commits:
>
> Commit: b4c6cd222f4789617f15fac8ddec6c7e8d6e9f99 (7.0.beta0 + 5 commits)
>
> while in the previous testing, it was correctly:
>
> Commit: c57069e9c66ad4f28038744f370fe6aac9c574d0 (7.4.beta4 + 9 commits)
>
> I cannot understand exactly how this happened...
>
>
>
>
>
> Frederic
>
> Le jeudi 15 septembre 2016 09:49:13 UTC+2, Kwankyu Lee a écrit :
> Dear all,
>
>
> A patchbot is supposed to run forever, but realistically I should stop it
> from time to time. Pushing ctrl-c stops the patchbot abruptly, and the
> subsequent run seems sometimes to show somewhat erroneous behavior, perhaps
> due to the spurious state of the files.
>
>
> Is there a way to stop the patchbot gracefully? Or am I worrying too much
> and ctrl-c is just ok?
>
>
> On Thursday, September 15, 2016 at 10:00:47 AM UTC+2, Frédéric Chapoton
> wrote:
>>
>> There is an option --count to say how many ticket to tests before to stop.
>> you may want to use that.
>>
>> You may even change it in your json config file during the patchbot run
>> (not tested if this works to stop the bot)
>
>
> Ok. I will try that. Thanks.
>
>>
>> But Ctrl-C should not be very problematic. Could you tell what kind of
>> problems you have seen ?
>
>
> I cannot say exactly. After a shutdown by ctrl-c, a subsequent doctesting
> used an old base:
>
> Commit: b4c6cd222f4789617f15fac8ddec6c7e8d6e9f99 (7.0.beta0 + 5 commits)
>
> while
>
>>
>>
>> Frederic
>>
>> Le jeudi 15 septembre 2016 09:49:13 UTC+2, Kwankyu Lee a écrit :
>>>
>>> Dear all,
>>>
>>> A patchbot is supposed to run forever, but realistically I should stop it
>>> from time to time. Pushing ctrl-c stops the patchbot abruptly, and the
>>> subsequent run seems sometimes to show somewhat erroneous behavior, perhaps
>>> due to the spurious state of the files.
>>>
>>> Is there a way to stop the patchbot gracefully? Or am I worrying too much
>>> and ctrl-c is just ok?
>
> --
> 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+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.

-- 
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+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Graceful shutdown of patchbot

[Kwankyu Lee]

> There is an option --count to say how many ticket to tests before to stop. 
> you may want to use that.
> You may even change it in your json config file during the patchbot run 
> (not tested if this works to stop the bot)


Ok. I will try that. Thanks.

But Ctrl-C should not be very problematic. Could you tell what kind of 
> problems you have seen ?


After a shutdown by ctrl-c, a subsequent doctesting used an old base and 
commits:

Commit: b4c6cd222f4789617f15fac8ddec6c7e8d6e9f99 (7.0.beta0 + 5 commits)

while in the previous testing, it was correctly:

Commit: c57069e9c66ad4f28038744f370fe6aac9c574d0 (7.4.beta4 + 9 commits)

I cannot understand exactly how this happened... 
  




Frederic

Le jeudi 15 septembre 2016 09:49:13 UTC+2, Kwankyu Lee a écrit :
Dear all,


A patchbot is supposed to run forever, but realistically I should stop it 
from time to time. Pushing ctrl-c stops the patchbot abruptly, and the 
subsequent run seems sometimes to show somewhat erroneous behavior, perhaps 
due to the spurious state of the files.


Is there a way to stop the patchbot gracefully? Or am I worrying too much 
and ctrl-c is just ok?


On Thursday, September 15, 2016 at 10:00:47 AM UTC+2, Frédéric Chapoton 
wrote:
>
> There is an option --count to say how many ticket to tests before to stop. 
> you may want to use that.
>
> You may even change it in your json config file during the patchbot run 
> (not tested if this works to stop the bot)
>

Ok. I will try that. Thanks.
 

> But Ctrl-C should not be very problematic. Could you tell what kind of 
> problems you have seen ?
>

I cannot say exactly. After a shutdown by ctrl-c, a subsequent doctesting 
used an old base:

Commit: b4c6cd222f4789617f15fac8ddec6c7e8d6e9f99 (7.0.beta0 + 5 commits)

while 
 

>
> Frederic
>
> Le jeudi 15 septembre 2016 09:49:13 UTC+2, Kwankyu Lee a écrit :
>>
>> Dear all,
>>
>> A patchbot is supposed to run forever, but realistically I should stop it 
>> from time to time. Pushing ctrl-c stops the patchbot abruptly, and the 
>> subsequent run seems sometimes to show somewhat erroneous behavior, perhaps 
>> due to the spurious state of the files.
>>
>> Is there a way to stop the patchbot gracefully? Or am I worrying too much 
>> and ctrl-c is just ok?
>>
>

-- 
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+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Graceful shutdown of patchbot

[Frédéric Chapoton]

There is an option --count to say how many ticket to tests before to stop. 
you may want to use that.

You may even change it in your json config file during the patchbot run 
(not tested if this works to stop the bot)

But Ctrl-C should not be very problematic. Could you tell what kind of 
problems you have seen ?

Frederic

Le jeudi 15 septembre 2016 09:49:13 UTC+2, Kwankyu Lee a écrit :
>
> Dear all,
>
> A patchbot is supposed to run forever, but realistically I should stop it 
> from time to time. Pushing ctrl-c stops the patchbot abruptly, and the 
> subsequent run seems sometimes to show somewhat erroneous behavior, perhaps 
> due to the spurious state of the files.
>
> Is there a way to stop the patchbot gracefully? Or am I worrying too much 
> and ctrl-c is just ok?
>

-- 
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+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.