Would anyone know about the possible cause for a higher IO Wait times being experienced by a RHEL 6.3 OS Server running on HP BL460c G7 Blade (With QMH2562 HBA) ? This server was recently connected to Brocade 5300 (on 6.4.0b firmware). We have noticed occasional loss of paths (through either of the 2 blade HBA ports). At times the server is spontaneously rebooting itself. Before the recent relocation work, this Blade Server was connected to HP 8/80 SAN Switch (on 6.4.0b firmware), at that time we had no such errors or higher IO wait times, but as soon we relocated the HP Blade Chassis to Brocade 5300 SAN switch we started experiencing higher IO Wait times and FC paths being dropped.
First off, FOS should be upgraded because the current release is pretty old.
Could you please post a diagram of the SAN topology? where the host is connected, its storage, the speed of each device, if there are ISLs, other switchs, etc.
Have you seen errors in any of the ports?
Also what interconnect modules are in the enclosure, virtual connect or actual switches, if the latter what brand,model,FW and mode are they running.
Perhaps related, so do you have a RCA for the intermittent reboot of the blade?
2 per haps related, are you running the HP fc enablement kit on OS or is this a vanilla RH with dev-mapper-multipath?