ct-grid

ct-grid@nikhef.nl

243 discussions

[Ct-grid] DMTCP anyone?
by Sander Klous 13 Jan '09

13 Jan '09

Hi, Ever used this? If so, does it work? > DMTCP is a distributed checkpointing system that can not only checkpoint > sequential programs, but also threaded programs (with pthreads), families > of processes (made with fork), and distributed processes across machines > (like MPI). You can find reference for it here: > > http://dmtcp.sourceforge.net/ > Would it be worthwhile to investigate DMTCP on grid and see how well it works by wrapping it around the job execution boundaries? Thanks, Sander

4 3

[Ct-grid] Grid Meetings in 2009
by Dennis van Dok 06 Jan '09

06 Jan '09

Hi *, in order to stay 'out of sync' with other regular meetings, the upcoming BiG Grid meetings are scheduled on the 1st and 3rd Mondays of each month. This should be easy enough to remember, and as a bonus the sirens at 1200h every first Monday of the month act as a timely warning to have an early lunch! To make things even easier, most of the meetings will be held in H3.31. Cheers, Dennis van Dok -- D.H. van Dok :: Software Engineer :: www.nikhef.nl :: www.vl-e.nl Phone +31 20 592 50 12 :: http://www.nikhef.nl/~dennisvd/

1 0

[Ct-grid] Happy new year: some NDPF stats of 2008 for entertainment
by David Groep 03 Jan '09

03 Jan '09

Hi all, A happy new year to all of you. Apart from all our regular accounting and billing, there are also a few nice graphs to extract from our accounting data. For the Nikhef part, a couple of nice figures for your enjoyment: In 2008, we processed 1441182 jobs, and on average 660 cores were in use averaged over the year. This is the equivalent of a little over 1 million days of work on a 1Ghz Pentium-III system, so if we had used pizza0-class boxes to do this work, we would have needed 1415 of them, or 70 racks full. The average CPU/wall efficiency was 86.3% December was the busiest month, with 221200 jobs, of which quite a sizeable faction (actually most) were submitted during the New-Year break. So, we can happily go on holidays, it's good for business :-) Also, 32 unique groups used our facility, from 25 different user communities (of which only 4 are dedicated to infrastructure testing). Many job submissions are also automated, or people like working over the week-end. Although working days see a slightly higher submission rate, it's not more than ~ 20% over the week-end submission rate. (see the graph below). And Wednesday is the most popular day to start working on the grid. In general, we see more submissions over the course of 2008, also prompted of course by the number of cores available. If you look at job submissions per core, it is rather constant (which means that in forst order all nodes we will buy and commissions will just be used. There is no slack in demand :-). Nice detail: if you reduce the capacity, like what happened to us is May/June 2008, the jobs keep coming, and thus job the submission count per core goes up. Enjoy the database! Cheers, DavidG. -- David Groep ** Nikhef, Dutch National Institute for Sub-atomic Physics,PDP/Grid group ** ** Room: H1.56 Phone: +31 20 5922179, PObox 41882, NL-1009DB Amsterdam NL **

1 0

[Ct-grid] next BiG Grid Meeting
by Dennis van Dok 19 Dec '08

19 Dec '08

Hi, the next BiG Grid meeting is *not* going to be on 22 December as was originally planned. I propose to start on the first Monday of the new year, so that will be 5 January, at 13.30h (location to be announced). If you intend to come, please let me know; if you have problems with this date and time, also please let me know because we need to stay 'out of sync' with other meetings that are regularly held Monday afternoons. Cheers, best wishes, have a safe New Year, Dennis van Dok -- D.H. van Dok :: Software Engineer :: www.nikhef.nl :: www.vl-e.nl Phone +31 20 592 50 12 :: http://www.nikhef.nl/~dennisvd/

3 3

[Ct-grid] Fwd: sage , a alternative for MAthematica
by Jeff Templon 10 Dec '08

10 Dec '08

fyi Begin forwarded message: > From: MArcel Vreeswijk <h73(a)nikhef.nl> > Date: 10 December 2008 09:46:10 GMT+01:00 > To: graven(a)nikhef.nl, Marcel Merk <marcel.merk(a)nikhef.nl>, atlas-pat(a)nikhef.nl > Subject: sage , a alternative for MAthematica > Reply-To: Marcel Vreeswijk <h73(a)nikhef.nl> > > > Hi, > From time to time, people need Mathematica which is not available > everywhere at Nikhef. > > Consider to use a very nice (open source) package: sage. See: http://www.sagenb.org/ > The possibilities are practically unlimted! > > you can run it on your own machine, but i created an account (<10 > seconds) and run it online. For example, if you forgot how to > integrate 1/(x^3+1) , just type: integral( 1/(x^3+1), x) and click > evaluate. (note that it uses python language). > > Have fun, MArcel >

1 0

[Ct-grid] Next BiG Grid meeting 8 December 13.00h
by Dennis van Dok 08 Dec '08

08 Dec '08

The next Big Grid meeting will be on Monday 8 December, from *13.00* to 14.00h in room N3.28. Note the earlier starting time! The agenda: - operations summary - middleware - applications - AOB Enjoy the weekend and/or pakjesavond, Dennis -- D.H. van Dok :: Software Engineer :: www.nikhef.nl :: www.vl-e.nl Phone +31 20 592 50 12 :: http://www.nikhef.nl/~dennisvd/

2 1

[Ct-grid] Next grid meeting Mon 8 Dec
by Dennis van Dok 02 Dec '08

02 Dec '08

Hi, The next BiG Grid Meeting is scheduled for Dec 8; the start time of this meeting is currently set to 13.30h. We are kindly requested to reschedule a little earlier (13.00-14.00) just this once, so another meeting can start on time in the same room. I'm inclined to consent unless there are reasonable objections. Please let me know. Thanks, Dennis -- D.H. van Dok :: Software Engineer :: www.nikhef.nl :: www.vl-e.nl Phone +31 20 592 50 12 :: http://www.nikhef.nl/~dennisvd/

1 0

[Ct-grid] Ganga 5.1.1 deployed
by Hurng-Chun Lee 25 Nov '08

25 Nov '08

Hi all, Just to announce that the latest stable release of Ganga (v 5.1.1) has been deployed at NIKHEF. It's available on NIKHEF desktops (e.g. elel20). - The release note: http://ganga.web.cern.ch/ganga/release/5.1.1/release/ReleaseNotes-5.1.1 - configuration change at NIKHEF: * job submission to LCG is adopting the recently installed gLite WMS service at NIKHEF To start using it at NIKHEF: http://www.nikhef.nl/pub/experiments/atlaswiki/index.php/Using_GANGA_with_A… comments and issue report: hn-atlas-dist-analysis-help(a)cern.ch Cheers, Hurng

1 0

[Ct-grid] Een idee voor ons nieuwste cluster?
by Tristan Suerink 24 Nov '08

24 Nov '08

http://tweakers.net/nieuws/56938/sgi-demonstreert-supercomputer-op-basis-va…

1 0

[Ct-grid] cannot start voms proxy on NIKHEF desktop
by Hurng-Chun Lee 24 Nov '08

24 Nov '08

hello, when doing "voms-proxy-init --vo atlas" on NIKHEF desktops (e.g. elel20), I got the following error: === [hclee@lxplus217:~/athena]$ cd .. GSS Major Status: Authentication Failed GSS Minor Status Error Chain: globus_gss_assist: Error during context initialization globus_gsi_callback_module: Could not verify credential globus_gsi_callback_module: Could not verify credential globus_gsi_callback_module: Invalid CRL: The available CRL has expired Trying next server for atlas. Creating temporary proxy ...................................... Done Contacting voms.cern.ch:15001 [/DC=ch/DC=cern/OU=computers/ CN=voms.cern.ch] "atlas" Failed Error: Could not establish authenticated connection with the server. GSS Major Status: Authentication Failed GSS Minor Status Error Chain: globus_gss_assist: Error during context initialization globus_gsi_callback_module: Could not verify credential globus_gsi_callback_module: Could not verify credential globus_gsi_callback_module: Invalid CRL: The available CRL has expired None of the contacted servers for atlas were capable of returning a valid AC for the user. === could you please have a look? Thanks! Cheers, Hurng

1 0

← Newer
1
...
16
17
18
19
20
21
22
...
25
Older →

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

ct-grid