Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On Mon, Jun 03, 2013 at 02:22:54PM +0200, Grzegorz Wolny wrote: > Hi all, > > I read article about DRBD 8.4.3 and performance improvement in > random tests: > http://blogs.linbit.com/p/469/843-random-writes-faster/ > > I compared DRBD 8.3.10 and 8.4.3 and achieved strange results. > > * 300GB volume synchronization time: > DRBD 8.3.10 - 00:18:24 > DRBD 8.4.3 - 00:52:06 > > * Sequence 300GB write time (DD) > DRBD 8.3.10 - 00:42:12 > DRBD 8.4.3 - 00:37:37 > > * Random write test time (FIO) > DRBD 8.3.10 - 03:19:07 > DRBD 8.4.3 - more than 1 day (after 3 hours interrupted) > > Last test confirms that DRBD 8.4.3 is better than 8.3.10 in random > test, but synchronization time is disturbing. > Why synchronization time is more than 2 times faster on DRBD 8.3.10? > What can I improve in my configuration? So you benchmark heavy IO during concurrent full resync. Not a particularly real-world benchmark, but whatever. Maybe this helps: http://lists.linbit.com/pipermail/drbd-user/2012-August/018900.html -- : Lars Ellenberg : LINBIT | Your Way to High Availability : DRBD/HA support and consulting http://www.linbit.com DRBD® and LINBIT® are registered trademarks of LINBIT, Austria. __ please don't Cc me, but send to list -- I'm subscribed