Hide Forgot
Description of problem: When there are too many unlink and create in the changelogm, the geo-rep status shows files synced in negative values, [root@redcell ~]# gluster v geo master 10.70.43.76::slave status detail MASTER NODE MASTER VOL MASTER BRICK SLAVE STATUS CHECKPOINT STATUS CRAWL STATUS FILES SYNCD FILES PENDING BYTES PENDING DELETES PENDING FILES SKIPPED ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- redcell.blr.redhat.com master /bricks/master_brick1 10.70.43.174::slave Active N/A Changelog Crawl -3002 0 0 0 3621 redlake.blr.redhat.com master /bricks/master_brick3 10.70.43.135::slave Active N/A Changelog Crawl -2781 0 0 0 3344 redcloak.blr.redhat.com master /bricks/master_brick2 10.70.43.76::slave Passive N/A N/A 0 0 0 0 0 redwood.blr.redhat.com master /bricks/master_brick4 10.70.42.151::slave Passive N/A N/A 0 0 0 0 0 Version-Release number of selected component (if applicable):glusterfs-3.4.0.45geo-1 How reproducible:Didn't try to reproduce, but it can be reproduced easily with steps given below Steps to Reproduce: 1.create and start a geo-rep relationship between master and slave 2. run following command on master mount point, ./crefi.py -n 100 --multi -b 10 -d 10 --random --max=2K --min=1K -t tar /mnt/master/test/ 3.wait for it to complete Actual results: the status detail shows negative values for files synced Expected results: It shouldn't show negative values in status detail. It is disastrous. Additional info:
files synced column is removed from status output in RHGS 3.1. If we introduce percistent store while working on RFE 988857, we can show number of Files synced. But with the existing limitation, this column is removed since it will mislead the user. Closing this bug for the same reason. Please reopen this bug if the issue is found again.