Re: [Condor-users] Questions about DAGMAN jobs & How to setup notification to other users in DAGMAN jobs?

["Alas, Alex [FEDI]" <aalas@xxxxxxxxxxxxx>]

Kent,
Thanks again for your response, I have pretty good idea of what to do
now. 
I have one more question, I noticed that when I submit my .dag file it
takes several minutes for the set of jobs A to start working in the
pool. Is that normal or if something related to my pool only? 
Concerning ideas on how to improve the documentation,... well, my basic
objection is that all the examples and tutorials are base for Unix\Linux
pools and there is very little or no reference for Windows Pools. I am
basically a Windows guy with very little knowledge of Linux\Unix, . If
you guys were to include example for windows pools with batch files,
etc;  I would have been easier to understand the application and how to
use it. Also I think in the manual there is too much emphasis on the
dependency features of the DAGMAN (I don't know why I got the impression
at first that that DAGMAN jobs were only if you needed to configure
dependency jobs) which I think is great and phenomenal but the fact the
DAGMAN can also be use for  non-dependency jobs is also very valuable to
make emphasis and for me was the main keystone feature I needed. 

I hope my input helps in something, definitely your has been very
valuable for my work.
Thanks again,
Alex 


-----Original Message-----
From: condor-users-bounces@xxxxxxxxxxx
[mailto:condor-users-bounces@xxxxxxxxxxx] On Behalf Of R. Kent Wenger
Sent: Thursday, July 23, 2009 12:55 PM
To: Condor-Users Mail List
Subject: Re: [Condor-users] Questions about DAGMAN jobs & How to setup
notification to other users in DAGMAN jobs?

On Wed, 22 Jul 2009, Alas, Alex [FEDI] wrote:

> Thanks for your answer Kent, I ran my dag job using your suggestion
and
> it seems to work fine. But I have an question here, you said in case I
> want to notify multiple users I need to create a final node that will
> send the notifications. How do you do that?

Just add another node to your DAG, and make that node a child of the
final 
node or nodes that you already have.  Then the "notification" node can 
just run a script that emails the notification to whoever you want.
(Note 
that this won't work quite the same as the normal Condor notification, 
because the email will only be sent if the DAG succeeds.)  (We've
actually 
talked about, but not yet implemented, an option to have a special 
"cleanup" node that gets run at the end if the DAG fails.  That would
make 
it possible to have multiple notifications sent if the DAG fails.)

> I am not a newbie in condor but I am not an expert neither and this
> DAGMAN topic is completely new to me. Maybe if you have some spare to
> time, would you be  kind enough to share that knowledge or maybe you
can
> point me to a document that explains how to do that?

Hopefully the above note makes sense...

> Concerning the question about the Rescue DAG, after reading the manual
> around 5 times I got it... thanks for your explanation as well on that
> toping

Hmm -- maybe that means we need to clarify that part of the manual.  I 
know it was one of the more difficult parts to write.  Any suggestions
on 
how to make this clearer?

Kent Wenger
Condor Team
_______________________________________________
Condor-users mailing list
To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx with
a
subject: Unsubscribe
You can also unsubscribe by visiting
https://lists.cs.wisc.edu/mailman/listinfo/condor-users

The archives can be found at: 
https://lists.cs.wisc.edu/archive/condor-users/