I would kindly ask someone if you could check my multipath.conf file, I have used this file for my SLES10SP3 Server with FCP attached LUNs, NetApp Metrocluster. But after takeover and giveback the filesystems created on these LUNs become read-only.
The part after device is looking like this
getuid_callout “/sbin/scsi_id -g -u -s /block/%n”
prio_callout “/sbin/mpath_prio_alua /dev/%n”
features “1 queue_if_no_path”
Thank you in advance for your tips and contributions.[/QUOTE]
As I normally do setups for EMC and Equallogic SAN’s… I don’t have personal experience with NetApp’s on SLES.
A quick look in the multipath.conf.defaults on a SLES 11 SP1 system I have running here shows some differences in multipath.conf section for the NETAPP:
I’m curious how you got to your multipath.conf (following a NetApp doc for example)? The path_checker parameter could be one to check (take care when changing these params with running production installations).
Note that the multipath.conf.defaults are normally a generic starting point and not always optimized… So always double check with the SAN vendor given guidelines and setup as the correct parameters also can depend on with settings have been configured within your SAN.
Also, could you post the output of ‘multipath -ll’?