Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On Fri, Jun 17, 2011 at 09:26:26AM +0200, motyllo8 wrote: > Hi, > > So, I am experimenting with handling I/O errors. > > My test scenario is following: > > I am writing some data, removing disk - io errors are then reported by kernel and handler is invoked (it is ok). > > I am not writing data, removing disk - io errors are reported by kernel but hanlder is not invoked immediately. No, I don't think there are "IO errors reported by kernel", at least no IO errors that are related to requests from DRBD. > So if I do not write any data to disk then I can not perform defined action exactly in time when failure occurs. What is reason of this? Obviously, you only realize that the car won't start once you try to use it. DRBD only knows about requests passing through it. If there are 0 requests passing through DRBD, all of these 0 requests will fail because of the broken/non-existent disk. And because 0 requests failed, 0 io-errors have been reported to DRBD, and 0 io-error handlers are invoked. Makes sense? -- : 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