Dear xrootd team,
Right now for ATLAS I am testing using Amazon EC2 and Elastic Block
Store (EBS) for xrootd storage cluster.
I am using 4 EC2 large instances (1 reserved and 3 spot instances) The 3
spot instances are data servers and
reserve instance is the redirector. Each dataset server has EBS device
attached for the storage.
I am using frm and xrdcp in extreme copy mode to stage files from US
ATLAS global xrootd storage (6 sites with the
files). I am seeing poor performance. The frm queues wedge and I have
to restart the frm_xfrd process to resume
the queues. Ratio of failed copies to sucessful copies is very large.
I should note that when I used the ephemoral storage on the instances I
did not see the same effect.
Are there any tunable timing parameters that might make the system more
forgiving given that it is running in a virturalized
environment. (Yes Andy.. I know ... I am like doing something the
system was not designed for)
Thanks,
Doug Benjamin
########################################################################
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
|