Re: [Alchemi-users] Slow Job Times

John Dehart Mon, 28 Aug 2006 14:08:57 -0700

Thanks Matt,

I'm still having the issue. It seems that the delay is nearly 5.1 minutes for each machine. I do have one machine that is working correctly (1min per run). I have checked the registry and other items and found nothing. If I watch the working directory in the machines I see the run work correctly and the results file will be sent back to the manager but it just sits for 5 minutes.

I have no idea why one machine would work and 7 others do not. Has anyone got a clue?

Thanks in advance,

John

----- Original Message ----
From: Matthew Riek <[EMAIL PROTECTED]>
To: John Dehart <[EMAIL PROTECTED]>
Sent: Thursday, August 24, 2006 11:05:01 PM
Subject: RE: [Alchemi-users] Slow Job Times

Hi John,

I tested Alchemi and had your exact problem not just between batches, but also when waiting for final jobs within a batch would cause 5-10 minute delays. All nodes were finished execution, but I would have a 5 - 10 minute delay while getting the serialized tasks back from the nodes through the manager to the user. The unusual thing is most GThreads finished, it's really just the last 1 or 2 that seem to have a huge pregnant pause before deciding to finish up.

I ended up having to write my own grid software as I couldn't find open source .NET based grid supporting file caching for both inputs and outputs of grid jobs. I am however still interested in knowing what this problem with alchemi is as I got a feeling no one else was experiencing it.

My tasks were requiring serialization of 3-4 megs per task, 80 PC cluster on a gigabit network.

Anyway, let me know if u have any luck with this.

-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of John Dehart
Sent: Friday, 25 August 2006 3:08 PM
To: [email protected]; [email protected]
Subject: [Alchemi-users] Slow Job Times

Guys,

First off let me say I'm new to the group and I have been working with alchemi for about 2 weeks now. Very impressed. I hope I can add to the progression.

Anyway, I have set up a test grid of 6 local machines. I am running batch jobs (multiple programs)with the alchemi_sub executable.

During my testing I am seeing the manger send jobs out with an approx. 5 minute delays between the completion of one set of jobs and the submital of the next. Is there something I should be doing to force the next set to move out more quickly?

Thanks in advance,

John

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

_______________________________________________
alchemi-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/alchemi-users

Re: [Alchemi-users] Slow Job Times

Reply via email to