Print

Print

@osschar Indeed LHCb reconstruction jobs have a different access pattern - they copy data directly to the WN and run off the local data. This is possible as we know exactly the size of the input data and have designed it to fit into the minimum storage available per job.

User / physics jobs stream data directly from the storage as the number of input files can be large (user dependent), totalling to 2 - 3 TB of data at times (or even more) - too big to fit into any local storage for the job.

The xrootd monitoring data coming to you are from the test system set up at RAL. These jobs exclusively run a minimal LHCb job which we created on request and reproduces the issue seen by user jobs. Hopefully fixing the problem with these jobs will solve this problem seen by all users (LHCb and possibly others).

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or unsubscribe.

[ { "@context": "http://schema.org", "@type": "EmailMessage", "potentialAction": { "@type": "ViewAction", "target": "https://github.com/xrootd/xrootd/issues/1259#issuecomment-673333031", "url": "https://github.com/xrootd/xrootd/issues/1259#issuecomment-673333031", "name": "View Issue" }, "description": "View this Issue on GitHub", "publisher": { "@type": "Organization", "name": "GitHub", "url": "https://github.com" } } ]

Use REPLY-ALL to reply to list

To unsubscribe from the XROOTD-DEV list, click the following link:
https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-DEV&A=1