[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

[dm-devel] Re: [PATCH][RFC] supporting different failover models



Joe Thornber wrote:
On Wed, Feb 11, 2004 at 03:34:23AM -0800, Mike Christie wrote:

It can be pretty high. I can get more detailed accounting of the costs in the morning.


That would be really useful, thanks.


Some devices/paths will show up as /dev/sda, but they can only accept basic commands like INQUIRY, other IO like reads and writes will just be failed. It really is vendor specific though. What exactly is the "driver for the path" is the question? I wanted to add the callouts to the "driver for the path", but I am not sure if this is the path-selector or what?


The path selector is there because people kept saying they wanted
different path selection policies for different hardware.  But since
there is now 1 path selector per group I can't see how it fits with
this problem.  Maybe we could add a couple of new methods to the path
selector to init/exit the group (yuck) ?

It becomes more fun when you consider that not every vendor will activate paths with the same command. Should they be able to reuse the RR ps, by having the command formatting somehow modular.



Thats what it looked like. Something that will ease this is being able to distinguish the error types hard, soft etc that way DM is not failing on any old error. I got no feedback when I posted, so Pat had been looking into this and I will soon.


We really need this. I had hoped for more feedback from the SCSI guys.


Are you referring to the work structure? Everything I added is just one per "struct multipath", and is allocated in the ctr.


I see that now.  Let's stick with the daemon approach for now,
while we're still working out _what_ this target should do.  You may
be right in the long run.

ok.





[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]