Parallel Workloads Archive: OSC Linux Cluster

The OSC Linux Cluster log

System: Linux cluster with 32 quad nodes and 25 dual nodes
Duration: Jan 2000 thru Nov 2001
Jobs: 80,714

This log contains about 22 months worth of data from accounting records produced by the Maui 3.0.7 scheduling software running on this cluster at Ohio Supercomputing Center (OSC).

The original log was provided online at Supercluster.org's HPC Wordload/Resource Trace Repository. In addition to the workload trace, a resource trace and documentation of the log's format were also provided (we have cached copies of these, see below). The extensive workload trace includes several fields which do not exist in the standard workload format, and hence were not converted. See the conversion notes for some caveats of the conversion to standard workload format.

Downloads:

OSC-Clust-2000-0 2.0 MB gz original log
OSC-Clust-2000-3.swf 1.1 MB gz converted log
OSC-Clust-2000-3.1-cln.swf 0.5 MB gz cleaned log -- RECOMMENDED, see usage notes
OSC-Clust-2000-1.swf 0.9 MB gz OLD VERSION of converted log (replaced 2 Aug 2006)
OSC-Clust-2000-2.swf 1.1 MB gz OLD VERSION of converted log (replaced 13 Nov 2011)
OSC-Clust-2000-2.1-cln.swf 0.5 MB gz OLD VERSION of cleaned log (replaced 13 Nov 2011)
(May need to click with right mouse button to save to disk)

Papers Using this Log:

This log was used in the following papers:

[feitelson04b] [sabin05] [talby07] [iosup08] [verma08] [aida09] [zakay13] [liang13] [zakay14] [feitelson14]

System Environment

This is a Linux cluster with two types of nodes: 32 quad-processor nodes, and 25 dual-processor nodes, for a total of 178 processors. The scheduling was done by Maui 3.0.7.

We have a cached copy of the resource trace which describes the nodes. The format can still be found at Adaptive Computing (or cached) -- need to scroll down.

Log Format

Full details were available from Supercluster.org, and can still be found at Adaptive Computing. We laso have a cached copy of this. However, the actual log deviates from this specification as indicated below.

Conversion Notes

The converted log is available as OSC-Clust-2000-3.swf. The conversion from the original format to SWF was done subject to the following. The conversion was done by a log-specific parser in conjunction with a more general converter module.

The differences between conversion 3 (reflected in OSC-Clust-2000-3.swf) and conversion 2 (OSC-Clust-2000-2.swf) is

The differences between conversion 2 (reflected in OSC-Clust-2000-2.swf) and conversion 1 (OSC-Clust-2000-1.swf) are

Usage Notes

The original log contains several flurries of very high activity by one specific user, which may not be representative of normal usage. These were removed in the cleaned version, and it is recommended that this version be used.
The cleaned log is available as OSC-Clust-2000-3.1-cln.swf.

A flurry is a burst of very high activity by a single user. The filter used to remove the flurries that were identified is

user=174 (44,617 jobs removed)
Note that the filters were applied to the original log, and unfiltered jobs remain untouched. As a result, in the filtered logs job numbering is not consecutive.

Further information on flurries and the justification for removing them can be found in:

The Log in Graphics

File OSC-Clust-2000-3.swf

weekly cycle daily cycle burstiness and active users job size and runtime histograms job size vs. runtime scatterplot utilization offered load performance

File OSC-Clust-2000-3.1-cln.swf (cleaned)

weekly cycle daily cycle burstiness and active users job size and runtime histograms job size vs. runtime scatterplot utilization offered load performance


Parallel Workloads Archive - Logs