instance file system will become read only after disabling a FC port

[KC Bi]

1. This bug is triggered will qualifying Canonical OpenStack(Mitaka) with EMC XtremIO;
2. The Ubuntu server deployed through MAAS is Ubuntu 14.04.5 LTS;
3. multipath-tools 0.4.9-3ubuntu7.13 is used;
4. Below multipath.conf is used:
defaults {
        verbosity 2
        wwids_file /etc/multipath/wwids
        user_friendly_names no
        flush_on_last_del yes
}
blacklist {
        devnode "^(ram|raw|loop|fd|md|dm-|sr|scd|st)[0-9]*"
        devnode "^hd[a-z]"
        devnode "^dcssblk[0-9]*"
        devnode "^cciss!c[0-9]d[0-9]*"
        devnode "^nvme.*"
        device {
                vendor "DGC"
                product "LUNZ"
        }
        device {
                vendor "EMC"
                product "LUNZ"
        }
        device {
                vendor "IBM"
                product "S/390.*"
        }
        device {
                vendor "IBM"
                product "S/390.*"
        }
        device {
                vendor "STK"
                product "Universal Xport"
        }
}
blacklist_exceptions {
}
devices {
        device {
                vendor XtremIO
                product XtremApp
                path_selector "queue-length 0"
                rr_min_io_rq 1
                path_grouping_policy multibus
                path_checker tur
                failback immediate
                fast_io_fail_tmo 15
        }
... other array's device definition ...
}
multipaths {
}

5. XtremIO array paths can be managed successfully:
ubuntu@CA31:~$ sudo multipath -ll
3514f0c5e4c0002aa dm-3 XtremIO ,XtremApp
size=5.0G features='0' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=1 status=active
| `- 13:0:7:1 sdq 65:0 active ready running
|-+- policy='round-robin 0' prio=1 status=enabled
| `- 13:0:6:1 sdr 65:16 active ready running
|-+- policy='round-robin 0' prio=1 status=enabled
| `- 12:0:6:1 sdt 65:48 active ready running
`-+- policy='round-robin 0' prio=1 status=enabled
  `- 12:0:7:1 sds 65:32 active ready running

6. Reproduce steps:
a. Launch a instance(cirros-0.3.4-x86_64-disk.img) with the help of OpenStack dashboard;
b. Attach the volume seen in step 5 to the instance;
c. Format and mount the volume in the instance, then start IO;
d. Disable a FC port(2 x paths will be impacted) at the switch side;
e. After a while(30 ~ 60 seconds), the file system based on the volume in the instance will become read only;
f. Restart the instance will clear the problem;
g. Based on the symptom, it seems multipath does not fail over IO of the instance to alive paths in time.

7. Logs from the host and instance are attached to this bug.

[KC Bi (bikecheng)]

After upgrading the OS of the array, the issue disappears.