[Sciserver-users] Compute node restart - Wed the 24.1. at 9:00!

[Jonas Haase]

Dear SciServer Users

The issue with the crashing kernel has been resolved. It turned out to be rather esoteric(*), but completely unrelated to the new compute images.
The newest ’SciServer’ images have been reinstated everywhere again, all newly started containers will contain updated software.

kind regards
Jonas


(*) esoteric details:
It turns out Jupyter notebooks really really don’t like it if there is a subdirectory called ‘code’ with a  __init__.py file in it.
This clashes with code.py from the Python standard library
https://docs.python.org/3/library/code.html
Please give you python modules a different descriptive name.


> On 25. Jan 2023, at 09:33, Jonas Haase <jhaase at mpe.mpg.de> wrote:
> 
> Dear all
> 
> Most compute nodes have been restarted. Unfortunately this has not resolved the issue with the unstable kernel in Jupyter Notebooks.
> I have therefore rolled the ’Sciserver’ compute image back to a older version where it seems to work while I try to figure out what is wrong.
> If you already have a container where things work as you want them to I would recommend to keep using that one for the time being.
> 
> Sorry for the inconvenience!
> 
> Jonas
> 
>> On 24. Jan 2023, at 14:26, Jonas Haase <jhaase at mpe.mpg.de <mailto:jhaase at mpe.mpg.de>> wrote:
>> 
>> Dear SciServer Users
>> 
>> We currently seem to have an issue with the memory on several compute nodes. The symptoms are crashing/restarting kernels in Jypyter notebooks and random segmentation faults running scripts.
>> Since this happens both with old and new containers my assumption is that the problem lies with the underlying compute nodes.
>> 
>> I would like to restart all compute nodes ** Tomorrow morning Wed the 24.1. at 9:00 **
>> Could I please ask you to finish your running scripts and save progress until then?
>> 
>> Apologies for the short notice.
>> 
>> kind regards
>> Jonas
>> 
>>> On 23. Jan 2023, at 16:49, Jonas Haase <jhaase at mpe.mpg.de <mailto:jhaase at mpe.mpg.de>> wrote:
>>> 
>>> Dear MPE SciServer Users
>>> 
>>> I would like to announce that I just have released a set of new Compute Images for the SciServer. 
>>> This means that every _new_ compute container you start from now on has a updated collection of software installed.
>>> 
>>> - The main change to the previous images is that the new baseline operating system is Ubuntu 20.04 instead of Ubuntu 18.
>>> - The installed Python version has been updated to 3.9 and R to version 4.2.1
>>> - The scientific python module collections have been updated to Anaconda 2022.10 and SciPy 1.9 respectively.
>>> - A large number of additional modules and software has been installed. You can see a partial list in the Readme file in 
>>> https://sciserver.mpe.mpg.de/dashboard/files/datavolumes/4 <https://sciserver.mpe.mpg.de/dashboard/files/datavolumes/4>
>>> and can see all installed python modules and their version with "conda list”
>>> - In the Xray images the HE software collections have been updated to HEASoft  6.31.1, CIAO 4.15, eSASS 20211214.0.5.
>>> - In addition the CIAO installation has been changed to be part of a conda environment instead of a separate software installation.
>>>  The Xray and eSASS images have additional buttons buttons and command line instructions to run with the CIAO kernel. 
>>>  Please note that this environment does not come with all the same modules installed as the base environment.
>>> - Lastly there is now a a easier way to modify your environment and have it stick between containers: In a interactive shell the bash setup will attempt to source the file 
>>>  Storage/[user]/persistent/bashrc if it exists. So you can add your preferred shell configuration to that file. 
>>> I hope you will find everything working as intended. If you encounter any issues, please let me know - It is at times challenging to get all the various software installations to play nicely with each other.
>>> 
>>> Kind regards
>>> Jonas
>>> 
>>> —
>>> Jonas Haase
>>> Max Planck Institute for Extraterrestrial Physics (MPE)
>>> Giessenbachstr. 1, 85748 Garching, Germany
>>> X5 1.3.08
>>> +49 89 30000 3706
>>> 
>>> -- 
>>> Sciserver-users mailing list
>>> Sciserver-users at lists.mpe.mpg.de <mailto:Sciserver-users at lists.mpe.mpg.de>
>>> https://lists.mpe.mpg.de/cgi-bin/mailman/listinfo/sciserver-users
>> 
> 
> —
> Jonas Haase
> Max Planck Institute for Extraterrestrial Physics (MPE)
> Giessenbachstr. 1, 85748 Garching, Germany
> X5 1.3.08
> +49 89 30000 3706
> 
> -- 
> Sciserver-users mailing list
> Sciserver-users at lists.mpe.mpg.de <mailto:Sciserver-users at lists.mpe.mpg.de>
> https://lists.mpe.mpg.de/cgi-bin/mailman/listinfo/sciserver-users <https://lists.mpe.mpg.de/cgi-bin/mailman/listinfo/sciserver-users>
—
Jonas Haase
Max Planck Institute for Extraterrestrial Physics (MPE)
Giessenbachstr. 1, 85748 Garching, Germany
X5 1.3.08
+49 89 30000 3706

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mpe.mpg.de/pipermail/sciserver-users/attachments/20230126/bd156deb/attachment.htm>