[dm-devel] multipathd is not detect restored fc-path
sda at rosenergo.com
sda at rosenergo.com
Sat Jul 15 08:13:19 UTC 2006
Hello!
I have multipathing configuration with path_grouping_policy = failover with two fibre channel simultaneous connections (active and standby). If I break active connection, multipathd detect this and switch data transfer to standby. But later, then connection restored, multipathd is not detect this fact automatically, command "multipath -l" show than old fibre channel remain failed. Multipathd will see that old channel is OK only if I manually remap paths with "multipath" command. This is not good, because if after then the first path broken and restored the second path fail too, conection to SAN will be completely unable to work. I suppose this is because of unproper configuration of multipathing.
Explain me, please, that is cause of this trouble and how I may correct it.
This is my configuration:
- disk storage - ds4800
- hba - two Emulex Corporation LP952 Fibre Channel Host Adapter
- OS - rhel as 4 update 3
- multipath toolkit source - device-mapper-multipath-0.4.5-12.0.RHEL4.i386.rpm
- multipathd config:
defaults {
multipath_tool "/sbin/multipath -v0"
udev_dir /dev
polling_interval 10
default_selector "round-robin 0"
default_path_grouping_policy failover
default_getuid_callout "/sbin/scsi_id -g -u -s /block/%n"
default_prio_callout "/bin/true"
default_features "0"
rr_wmin_io 100
failback immediate
user_friendly_names yes
}
multipaths {
multipath {
wwid 3600a0b8000115280000040b9448f8cea
alias mp03
path_grouping_policy failover
path_selector "round-robin 0"
failback immediate
}
multipath {
wwid 3600a0b8000115280000040cd448f8d1e
alias mp04
path_grouping_policy failover
path_selector "round-robin 0"
failback immediate
}
}
devices {
device {
vendor "IBM"
product "1815 FAStT"
path_grouping_policy failover
path_checker tur
}
}
Thanks,
Dmitry Sedov, NP ATS.
More information about the dm-devel
mailing list