Bug #2907
rados benchmarking tool which does not always do creates
% Done:
0%
Source:
Development
Tags:
Backport:
Regression:
No
Severity:
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):
Description
Features:
Pluggable distribution for choosing objects (zipifan?, random?, sequential?)
configurable number of objects
configurable object size
configurable io size distribution
configurable io offset distribution
configurable read/write split (0-1)
configurable num in flight ops
latency bin size
Reports:
each second avg throughput
each second avg iops
each second 50,90,99% latency
overall avg for above
History
#1 Updated by Josh Durgin over 11 years ago
Why not record the raw data and let other tools produce percentiles and other statistics?
#2 Updated by Sage Weil about 11 years ago
- Status changed from New to Resolved