On Sat, Feb 21, 2004 at 12:40:10AM +0100, Lars Marowsky-Bree wrote: > On 2004-02-20T17:40:42, > Heinz Mauelshagen <Mauelshagen@xxxxxxxxxx> said: > > > > > in a fabric, which come back online shortly later when all paths have been > > > > failed by the multipath target. > > > You think EIO is the wrong response to this ? > > In case all paths come back after a reasonably short time it is. > > Can you define 'reasonably short time'? Defined by the amount it takes for a full system reboot and application startup, which can take multiple minutes. IOW: defined by the user. > > If they come back faster than a single SCSI timeout: we are covered. Of course. > > Otherwise, I'd say we should place the limit at 'We can only queue so > much until we run out of memory'; alas, the failure case which you bring > up is swapping to a fully disconnected m-p device under OOM, and that's > pathological. Is it ? If user has a bullet prove multipath solution in operation, which needs to be able to reuse a repaired path after the system went OOM, he'll surely prefer this vs. a longish system reboot with client recovery (thise might be thousands) including lots of helpdesk inquiries. We should give user the freedom of choice, if he prefers to take a system starvation for a definable amount of time -or- wants to fail the filesystem/application directly after all paths went south. I had system environments where a full system reboot took ~15 minutes. In such cases you'll surely would appreciate to have the option to repair a path (eg, replacing the dead FC cable) in order to recover sooner. > > > Sincerely, > Lars Marowsky-Brée <lmb@xxxxxxx> > > -- > High Availability & Clustering \ ever tried. ever failed. no matter. > SUSE Labs | try again. fail again. fail better. > Research & Development, SUSE LINUX AG \ -- Samuel Beckett > > -- > > dm-devel@xxxxxxxxxx > https://www.redhat.com/mailman/listinfo/dm-devel -- Regards, Heinz -- The LVM Guy -- *** Software bugs are stupid. Nevertheless it needs not so stupid people to solve them *** =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Heinz Mauelshagen Red Hat GmbH Consulting Development Engineer Am Sonnenhang 11 56242 Marienrachdorf Germany Mauelshagen@xxxxxxxxxx +49 2626 141200 FAX 924446 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-