Re: [HTCondor-devel] Contributing to HTCondor


Date: Fri, 15 Dec 2017 01:44:30 +0100
From: Oliver Freyermuth <freyermuth@xxxxxxxxxxxxxxxxxx>
Subject: Re: [HTCondor-devel] Contributing to HTCondor
Am 13.12.2017 um 22:05 schrieb Greg Thain:
> Dear Oliver:
> 
> I apologize for the delay in getting back to you and the problems with email.  We are interested in contributions in general and yours in particular.
> 
> In particular, we have some ideas for how to implement condor_ssh_to_job for various container technologies, perhaps it would be best to  talk interactively about this -- would you be available next week for a skype or other voip chat?
> 
> -greg thain
Dear Greg,

thanks for the positive reply! 

Yes, an interactive talk might be best. My colleague Peter Wienemann here in Bonn (Germany) is in CC (and also on the list),
we are working together on our HTCondor installation and the integration with containers, so we would both like to join the discussion if this is fine with you. 

From our side, Monday or Wednesday next week would be best. Since we are in "UTC+1" here, I would propose something like 5pm in our timezone (i.e. 4pm UTC),
which would be 10am for you. We are usually using Vidyo (due to our affiliation with CERN) which offers a WebRTC client
and can be used without account (as long as the meeting itself is created by the owner of a CERN account). Otherwise, Skype would also be possible. 

Cheers and all the best,
Oliver

> 
>>
>> A discussion on how to implement the "container universe" in general (to overcome the current issues, e.g. allow running sshd outside of the container,
>> and enter it unprivileged) would also be helpful, unless this is already on the schedule of development. 
>>
>> Many thanks and all the best,
>> Oliver
>>
>>
>> Am 13.12.2017 um 17:54 schrieb Tim Theisen:
>>> Hello Oliver,
>>>
>>> Your work is of interest to us and one our developers has tried to contact you via email. Unfortunately, the message bounced for some reason. Someone will reach out to you very soon.
>>>
>>> We do welcome contributions. I apologize for our lack of response so far.
>>>
>>> ...Tim
>>>
>>>
>>> On 12/12/2017 06:23 PM, Oliver Freyermuth wrote:
>>>> Dear developers, 
>>>>
>>>> since I have not yet received any feedback and from the archives it seems htcondor-devel has not seen any bidirectional activity
>>>> in the last months, I wonder whether this list is actually dead? 
>>>>
>>>> Receiving no feedback, I have in the meantime opened a PR on GitHub:
>>>> https://github.com/htcondor/htcondor/pull/18
>>>> Sadly, this is now also over a month old, by now has conflicts, and I did not receive any feedback on it. 
>>>>
>>>> I also asked related questions and given suggestions concerning the container implementations in HTCondor on htcondor-users,
>>>> again without reply (not unexpected, since I was asking in a conceptual way, for which the devel-list seems better suited). 
>>>>
>>>> So it seems to me all channels which I would expect to be used for interactions with developers are dead. 
>>>> Are external contributions welcome at all? 
>>>>
>>>> Cheers,
>>>> Oliver
>>>>
>>>> Am 02.11.2017 um 18:06 schrieb Oliver Freyermuth:
>>>>> Dear developers,
>>>>>
>>>>> since my last question might have been too specialized for others to find (especially in htcondor-users),
>>>>> I'm now asking in a more generic manner on the devel mailing list. 
>>>>>
>>>>> What's the best way to get started with HTCondor development, and in which way are contributions expected? 
>>>>>
>>>>> I would be interested in writing a charliecloud "plugin" (very much like the singularity plugin, only the wrapper and of course some parameters
>>>>> are different / do not exist). 
>>>>> Looking at the way the singularity plugin implementation is done, the procedure seems straightforward. 
>>>>>
>>>>> Background and reasoning (already explained on the htcondor-users list):
>>>>> Charliecloud should allow to easily resolve the condor_ssh_to_job issue,
>>>>> since an unprivileged "nsenter" call is sufficient to enter the container. This could be done via the authorized-keys injection or ForceCommand. 
>>>>> So as Cluster-admin it would be my first choice in terms of security (no setuid root / no root daemon) whenever user namespaces are sufficient,
>>>>> especially since it can run extracted containers built with docker or singularity, so no new work is needed. 
>>>>> That's why I'd like to see it supported in HTCondor, but of course nothing comes for free, so I'm offering to implement things. 
>>>>>
>>>>> Do I simply get started, at some point submit a PR on github / a mail here, and then we discuss implementation details? 
>>>>> Or is there another preferred way? 
>>>>>
>>>>> Or is somebody else interested in it / already working on that? 
>>>>>
>>>>> Cheers,
>>>>> Oliver
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> HTCondor-devel mailing list
>>>>> HTCondor-devel@xxxxxxxxxxx
>>>>> https://lists.cs.wisc.edu/mailman/listinfo/htcondor-devel
>>>>>
>>>>
>>>> _______________________________________________
>>>> HTCondor-devel mailing list
>>>> HTCondor-devel@xxxxxxxxxxx
>>>> https://lists.cs.wisc.edu/mailman/listinfo/htcondor-devel
>>> -- 
>>> Tim Theisen
>>> Release Manager
>>> HTCondor & Open Science Grid
>>> Center for High Throughput Computing
>>> Department of Computer Sciences
>>> University of Wisconsin - Madison
>>> 4261 Computer Sciences and Statistics
>>> 1210 W Dayton St
>>> Madison, WI 53706-1685
>>> +1 608 265 5736
>>>
>>
>>
>>
>> _______________________________________________
>> HTCondor-devel mailing list
>> HTCondor-devel@xxxxxxxxxxx
>> https://lists.cs.wisc.edu/mailman/listinfo/htcondor-devel
> 
> 
> 
> _______________________________________________
> HTCondor-devel mailing list
> HTCondor-devel@xxxxxxxxxxx
> https://lists.cs.wisc.edu/mailman/listinfo/htcondor-devel
> 


-- 
Oliver Freyermuth
Universität Bonn
Physikalisches Institut, Raum 1.047
Nußallee 12
53115 Bonn
--
Tel.: +49 228 73 2367
Fax:  +49 228 73 7869
--

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

[← Prev in Thread] Current Thread [Next in Thread→]