Hi Vlad, This is a two-step process.  First, you need to get the information into the schedd ad.  For this, look at the parameter: SCHEDD_COLLECT_STATS_FOR_<Name> This will create a statistics collection for all jobs matching that _expression_. Next, you will want to create a new Ganglia metric for the statistic - you can look at the example of running jobs per schedd as an inspiration. Let me know if you need more pointers and I can provide them when back at my laptop. Brian Sent from my iPhone 
  |