Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
Hi, we have 2 systems with DRBD. The DRBD sync is not active at the moment, because we are not able to resync the devices again. There is heavy traffic on the master via samba and nfs. The slave has about 350G data left to sync. If we activate the DRBD daemon on the slave and the traffic on the master grows too much, the file system on the master will not response anymore. A simple "ls" lasts 30+ seconds. About 1 minute later everything is normal again. Within these minute the DRBD sync speed is normal or even rises to 10-20Mb/s. The syncer rate is set to 5MB/s, but lowering it to 500kb/s does not make any difference. First we thought it’s a problem with our raid controllers (http://forums.storagereview.net/index.php?showtopic=25923 and http://community.novacaster.com/showarticle.pl?id=7395&n=8001), but the sync speed should also be zero or at least very low while these gaps, but it is not. The situation is completely reproducible. Certain jobs produces these gaps every time they run. specs (identical systems): - Xeon 3,0 Ghz - Memory 2GB - 3ware 9500 - 1x Raid 5 with 5x250G - 1x Raid 5 with 5x500G - 1gb cross over connection - Debian Etch, 2.6.18-4-amd64 - version: 0.7.21 (api:79/proto:74) Thanks, Arne