05-08-2013 07:43 PM
Hello all,
i would like to replace my failed disk via the vxdiskadm utility
05-27-2013 09:52 PM
Hi KJBSS,
How do i check what you mean by this
"Do you have this drive on an array that has active/passive support to your host? "
Im not really that familiar with symantec
05-28-2013 07:01 AM
Here's an example of a disk device that is on an active/passive array:
Use vxdisk list on the device in order to see the paths to the device and whether they are all primary or not:
$ vxdisk list c1t5006016041E0B566d26s2
Device: c1t5006016041E0B566d26s2
devicetag: c1t5006016041E0B566d26
type: auto
hostid:
disk: name= id=1228990244.166.myhost81
group: name=paytdg id=1228990226.158.myhost81
info: format=cdsdisk,privoffset=256,pubslice=2,privslice=2
flags: online ready private autoconfig
pubpaths: block=/dev/vx/dmp/c1t5006016041E0B566d26s2 char=/dev/vx/rdmp/c1t5006016041E0B566d26s2
guid: {f8ee5150-1dd1-11b2-a075-00144fb7bfa0}
udid: DGC%5FRAID%205%5F6ARA600581%5F6006016077002100249BE954D575DD11
site: -
version: 3.1
iosize: min=512 (bytes) max=2048 (blocks)
public: slice=2 offset=65792 len=2031232 disk_offset=0
private: slice=2 offset=256 len=65536 disk_offset=0
update: time=1368695151 seqno=0.45
ssb: actual_seqno=0.2
headers: 0 240
configs: count=1 len=48144
logs: count=1 len=7296
Defined regions:
config priv 000048-000239[000192]: copy=01 offset=000000 enabled
config priv 000256-048207[047952]: copy=01 offset=000192 enabled
log priv 048208-055503[007296]: copy=01 offset=000000 enabled
lockrgn priv 055504-055647[000144]: part=00 offset=000000
Multipathing information:
numpaths: 4
c1t5006016041E0B566d26s2 state=enabled type=primary
c3t5006016141E0B566d26s2 state=enabled type=primary
c3t5006016941E0B566d26s2 state=enabled type=secondary
c1t5006016841E0B566d26s2 state=enabled type=secondary
You can see above that some paths to the same LUN-device are "type=secondary" and others are "type=primary". You cannot access LUN-devices via the secondary paths. However, you should note that the "pubpaths: block=/dev/vx/dmp/c1t5006016041E0B566d26s2 char=/dev/vx/rdmp/c1t5006016041E0B566d26s2" line shows you how you can realiably access the LUN no matter which paths to the LUN are currently "primary". I:
$ prtvtoc /dev/vx/dmp/c1t5006016041E0B566d26s2
* /dev/vx/rdmp/c1t5006016041E0B566d26s2 partition map
*
* Dimensions:
* 512 bytes/sector
* 16 sectors/track
* 4 tracks/cylinder
* 64 sectors/cylinder
* 32768 cylinders
* 32766 accessible cylinders
*
* Flags:
* 1: unmountable
* 10: read-only
*
* First Sector Last
* Partition Tag Flags Sector Count Sector Mount Directory
2 5 01 0 2097024 2097023
7 15 01 0 2097024 2097023
Below you can see the error you can get if you try to access the device via a non-primary path:
$ prtvtoc /dev/dsk/c1t5006016841E0B566d26s2
prtvtoc: /dev/rdsk/c1t5006016841E0B566d26s2: Unable to read Disk geometry errno = 0x5
Below you can see that there is no error when you access the device via a primary path:
$ prtvtoc /dev/dsk/c1t5006016041E0B566d26s2
* /dev/dsk/c1t5006016041E0B566d26s2 partition map
*
* Dimensions:
* 512 bytes/sector
* 16 sectors/track
* 4 tracks/cylinder
* 64 sectors/cylinder
* 32768 cylinders
* 32766 accessible cylinders
*
* Flags:
* 1: unmountable
* 10: read-only
*
* First Sector Last
* Partition Tag Flags Sector Count Sector Mount Directory
2 5 01 0 2097024 2097023
7 15 01 0 2097024 2097023
Below you can see your controllers to your devices and their enclosure type:
$ vxdmpadm listctlr all
CTLR-NAME ENCLR-TYPE STATE ENCLR-NAME
=====================================================
c0 Disk ENABLED disk
c1 EMC_CLARiiON ENABLED emc_clariion0
c3 EMC_CLARiiON ENABLED emc_clariion0
c1 EMC_CLARiiON ENABLED emc_clariion1
c3 EMC_CLARiiON ENABLED emc_clariion1
EMC_CLARiiON arrays are "Active/Passiv" arrays, where only certain paths are active at a time.
You can use the "vxdmpadm listctlr all" command on your system to see what kind of arrays you have, if any, in your configuration.
Hope that helps, but I doubt it, because you have an I/O error, which is much more a hardware-physical layer problem. You need your hardware looked at by your array's hardware Field Engineer (or possibly your server's hardware engineer if the problem is with the HBA in your host).
05-29-2013 12:21 AM
Hi Everyone,
thank you for the answers :) its starting to make sense now.
After all the research, is there a way to enable the path of this disk?
05-29-2013 03:23 AM
What was the resolution to your earlier I/O errors?
If there are no known media nor SAN issues, Dynamic Multi Pathing (DMP) disables/enables disk paths after receiving SCSI Errors can be caused if your disk has no valid disk label on it (from TECH188277):
On the solaris server there were many disks with no disk label. DMP was unable to check the disk's associated paths because of there being no VTOC, so ended up failing the device and associated paths. Subsequently, DMP re-enables the path and disables it once again at the next check, and so on.
To enable a path:
vxdmpadm enable ctlr=
...where the CTLR-NAME and ENCLR-NAME are displayed via the 'vxdmpadm listctlr all' command.
Make sure you label c0t2d0s2:
$ format c0t2d0 > label > y > q
All the while, monitor your OS and DMP logs for errors. IE:
tail -f /var/adm/messages tail -f /var/adm/vx/dmpevents.log
06-21-2013 03:44 AM
chuckchang23,
As others have pointed out (eg: kjbss, Gaurav), it appears there are problems with access / i/o to the disk - as you are unable to run prtvtoc at times but other times it works.
From the output provided, it appears the disk is local / not SAN or EMC powerdevice (as the other powerdevices are showing correctly, and vxdisk list shows it's an LSI disk).
Check that it appears correctly in format, if there are errors shown in iostat -En, the status in cfgadm.
To reenable the disk in vxvm, it needs to be accessible (ie: you need to sort out the i/o errors), and to have a valid label (ie: if you have intermittent problems running prtvtoc to display the label, then vxvm is going to show the disk in error as it won't be able to read the label either).
Also note per Yasuhisa's comment ( https://www-secure.symantec.com/connect/forums/vxdiskadm-replace-disk-does-not-show-my-failed-disk#comment-8727391 ) - it appears this disk has been used outside vxvm, as your prtvtoc shows s3 and s6 were previously mounted. So whatever is using the disk at the moment may have had its own reasons for removing/not having slice 2 -- so creating s2 now may cause issues.
If you want to use the disk in vxvm, you will need to relabel the disk; however unless you know the history of the machine / have confirmed whatever was using it previously is definitely not impacted by this, you should exercise caution / be prepared to potentially restore a backup of any data that was on this disk as the relabel may render the previous data inaccessible.
regards,
Grace
07-03-2013 03:38 AM
Hi chuckchang23,
Looks like you are making some progress thanks to the good replies here. Would be best to progress this type of enquiry over in the Storage Foundation forum as this not really a cluster server related command.
Cheers