This contribution describes a number of actions and novel solutions introduced by the Spanish CMS community in order to facilitate the inclusion of BSC resources into the CMS computing infrastructure for their use by the collaboration. In particular, new HTCondor developments that allow condor daemons to interact via shared filesystems.
The CMS Computing Submission Infrastructure group manages and exploits a set of HTCondor pools to satisfy the experiment computing needs. The biggest of those pools, the so-called CMS Global Pool, currently aggregates nearly 400k CPU cores dynamically from pledged and opportunistic WLCG resources.
Historically, authentication among the diverse components of the infrastructure used the Grid...
To complement grid production workflows, the National Analysis Facility (NAF) at DESY aims to provide individual users with an easy way without hurdles to process their data fast.
Since users are used to plain file paths, a number of storages are mounted as network mounts on the NAF HTCondor nodes. But with users showing a wide range of access patterns, we encountered occassionally problems...
An Idiosyncratic Airing of Grievances, Condor-Style
At final we have a large unifed cluster.
4 years ago there were three clusters sharing the same infrastructure (network and data center).
A dedicated cluster per team.
We decided to join forces and recreate a large cluster to better utilize the resources.
Docker volumes able us to share compute resources and allow different storage access per team.
In addition for that we are...
I will explain how we uses the functionality in the HTCondor Start Daemon to manage and monitor our batch worker nodes
