Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On Tue, Feb 26, 2008 at 02:21:31PM +0100, Andreas Haase wrote: > Lars, > > Am Dienstag, den 26.02.2008, 14:04 +0100 schrieb Lars Ellenberg: > > if the other clusters are fine, what is the difference, > > what makes this cluster behave "bad"? > > it is one of the first clusters using this combination of kernel and > drbd version. We have at least one other cluster, that does not have > this problem. But that is completely different hardware/controller. > > Do you have any idea, why the kernel reponds with an unexpected error > code? no. but maybe I just misunderstand the internal kernal api, and we should interpret EIO in response to blkdev_issue_flush as equivalent to EOPNOTSUPP. > Or are there plans to introduce a rate limit for such error > messages? sure. probably we even need to make it configurable, so you could switch it off completely. drbd would then never even try to flush. -- : Lars Ellenberg http://www.linbit.com : : DRBD/HA support and consulting sales at linbit.com : : LINBIT Information Technologies GmbH Tel +43-1-8178292-0 : : Vivenotgasse 48, A-1120 Vienna/Europe Fax +43-1-8178292-82 : __ please use the "List-Reply" function of your email client.