The profiler_agent is scanning our EMC gatekeeper LUNs as part of the disk performance polling. These polls are generating errors in the syslog that we would like to prevent. Is it possible to exclude a disk from scanning by device name for a specific server? I've seen quite a few posts on how to filter the data out, but that of course is after the data is pulled from the clients and in our case, generating the errors. We don't want to turn off the disk performance polling completely, as we do want the data for valid devices. Anyone know if the exclude at the client level is possible?
Thanks,
Ameer
Error example:
Buffer I/O error on device sden, logical block 0
end_request: I/O error, dev sden, sector 2097144
Buffer I/O error on device sden, logical block 262143
end_request: I/O error, dev sden, sector 2097144
Buffer I/O error on device sden, logical block 262143
end_request: I/O error, dev sden, sector 0
Buffer I/O error on device sden, logical block 0
end_request: I/O error, dev sden, sector 0
end_request: I/O error, dev sdfz, sector 0
end_request: I/O error, dev sdfz, sector 0
end_request: I/O error, dev sdfz, sector 0
end_request: I/O error, dev sdfz, sector 0
end_request: I/O error, dev sdfz, sector 2097144
end_request: I/O error, dev sdfz, sector 2097144
end_request: I/O error, dev sdfz, sector 0
end_request: I/O error, dev sdfz, sector 0
