Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On 06/21/2011 09:51 PM, wang xuchen wrote: > Digimer, > > Thanks for your reply. > > I come up with the number 300M from DRBD official website "A good rule > of thumb for this value(rate parameter) is to use about 30% of the > available replication bandwidth. ". I use 10G Ethernet card for > replication traffic, 1024M * 0.3 is 300M. What wrong with my calculation? > > Ben That is your network speed. Have you calculated your maximum disk I/O? During resync, you should see how quickly the resync is running at. Is it actually going at ~300MB/sec? If your IO is well over 300MB/sec, then I am not sure off hand what is causing the slowdown. Are all ten DRBD resources on the same set of drives? If so, then perhaps the disks' read/write heads are flogged? To test, can you 'pause-sync' on all but one resource at a time and see if performance returns? If it does, then resume-sync one resource at a time and see at what point the performance becomes unacceptable. -- Digimer E-Mail: digimer at alteeve.com Freenode handle: digimer Papers and Projects: http://alteeve.com Node Assassin: http://nodeassassin.org "I feel confined, only free to expand myself within boundaries."