Login or Register to Ask a Question and Join Our Community


Solaris

Mpathadm showing no paths to 1 lun, others luns are fine

Login to Discuss or Reply to this Discussion in Our Community

 
Thread Tools Search this Thread
Operating Systems Solaris Mpathadm showing no paths to 1 lun, others luns are fine
# 1  
Old 10-31-2013
Mpathadm showing no paths to 1 lun, others luns are fine
Hi,

I've noticed that mpathadm states that one of our luns has no active paths:

Code:
       /dev/rdsk/c6t60000970000298700009533031324333d0s2
                Total Path Count: 4
                Operational Path Count: 4
        /dev/rdsk/c6t60000970000298700009533031333037d0s2
                Total Path Count: 4
                Operational Path Count: 0
        /dev/rdsk/c6t60000970000298700009533031324634d0s2
                Total Path Count: 4
                Operational Path Count: 4

The same lun shows up in iostat:

Code:
c6t60000970000298700009533031333037d0 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0
Vendor: EMC      Product: SYMMETRIX        Revision: 5874 Serial No:
Size: 103.08GB <103080591360 bytes>
Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0
Illegal Request: 2 Predictive Failure Analysis: 0

And is mirror in a zpool:

Code:
  pool: export
 state: ONLINE
 scrub: resilver completed after 0h2m with 0 errors on Wed Oct 30 14:42:42 2013
config:

        NAME                                       STATE     READ WRITE CKSUM
        export                                     ONLINE       0     0     0
          mirror-0                                 ONLINE       0     0     0
            c6t6001438002A57F6C0000800009B60000d0  ONLINE       0     0     0
            c6t60000970000298700009533031333037d0  ONLINE       0     0     0  36.4G resilvered

If the lun had no active paths, surely zfs would have complained? I only just added this lun to the zpool recently too. Is this a reporting anomoly - or deeper issue here?

Code:
root@server02 # mpathadm show lu /dev/rdsk/c6t60000970000298700009533031333037d0s2
Logical Unit:  /dev/rdsk/c6t60000970000298700009533031333037d0s2
        mpath-support:  libmpscsi_vhci.so
        Vendor:  EMC
        Product:  SYMMETRIX
        Revision:  5874
        Name Type:  unknown type
        Name:  60000970000298700009533031333037
        Asymmetric:  no
        Current Load Balance:  round-robin
        Logical Unit Group ID:  NA
        Auto Failback:  on
        Auto Probing:  NA

        Paths:
                Initiator Port Name:  10000000c978d125
                Target Port Name:  50000975f0002581
                Override Path:  NA
                Path State:  unavailable
                Disabled:  no

                Initiator Port Name:  10000000c978d0c3
                Target Port Name:  50000975f0002589
                Override Path:  NA
                Path State:  unavailable
                Disabled:  no

                Initiator Port Name:  10000000c978d0c3
                Target Port Name:  50000975f0002585
                Override Path:  NA
                Path State:  unavailable
                Disabled:  no

                Initiator Port Name:  10000000c978d125
                Target Port Name:  50000975f000258d
                Override Path:  NA
                Path State:  unavailable
                Disabled:  no

        Target Ports:
                Name:  50000975f0002585
                Relative ID:  0

                Name:  50000975f0002589
                Relative ID:  0

                Name:  50000975f0002581
                Relative ID:  0

                Name:  50000975f000258d
                Relative ID:  0

---------- Post updated at 04:03 AM ---------- Previous update was at 03:28 AM ----------

the plot thickens, after checking messages, I found the following. And the lun paths are offline/onlining constantly.

Code:
WARNING: kstat_create('mdi', 0, 'ssd301.t4.fp1'): namespace collisi
on
Oct 30 14:35:39 ldn-ux-kon02 genunix: [ID 834635 kern.info] /scsi_vhci/ssd@g60000970000298700009533031333037 (ssd301) multipath
 status: optimal, path /pci@3,700000/SUNW,emlxs@0,1/fp@0,0 (fp1) to target address: w50000975f0002585,4 is online Load balancin
g: round-robin

googling does not reaveal much. I'll see if I can log a case with Oracle.
Last edited by badoshi; 11-01-2013 at 06:35 AM..
# 2  
Old 10-31-2013
Hi badoshi,

Last time I had a similar issue, we fixed it by issuing a forcelip via luxadm. Some might look at it as if it was a demon but it helped me in the past.

To list the ports:
Code:
> root@oabamacare1:~# luxadm -e port
/devices/pci@0,0/pci8086,340c@5/pci10df,fc42@0/fp@0,0:devctl       CONNECTED
/devices/pci@0,0/pci8086,340c@5/pci10df,fc42@0,1/fp@0,0:devctl     NOT CONNECTED
/devices/pci@0,0/pci8086,340e@7/pci111d,806e@0/pci111d,806e@4/pci10df,fc42@0/fp@0,0:devctl  CONNECTED
/devices/pci@0,0/pci8086,340e@7/pci111d,806e@0/pci111d,806e@4/pci10df,fc42@0,1/fp@0,0:devctl  NOT CONNECTED

Then reinitialize the one giving you headache via -e forcelip
Code:
> root@obamacare1:~# luxadm -e forcelip /devices/pci@0,0/pci8086,340c@5/pci10df,fc42@0/fp@0,0:devctl

But be careful:
Code:
     -e forcelip enclosure[,dev] ... |  pathname...

         Force the link to reinitialize, using the Loop Initiali-
         zation  Primitive (LIP) sequence. The enclosure or path-
         name can specify any device on the loop. Use  the  path-
         name to specify a specific path for multiple loop confi-
         gurations.

         This is an expert only command and should be  used  with
         caution. It will reset all ports on the loop.

Might be worth opening a case with Oracle. Let us know what they recommend or if you give my idea a shot.
# 3  
Old 10-31-2013
This sure looks like some sort of bug.

If no paths were present for device, device would be offline in format, and you would not be able to manipulate it at all. Zpool status would be degraded.

Can you show output of other working disk from that same storage using same paths via mpathadm show lu.
If it's ok for other working disk, report it.

You can try forcelip, but do it one fabric at the time, monitoring syslog for return on paths before issuing it on next.
# 4  
Old 11-01-2013
I've logged a ticket with Oracle about this issue, and to their credit were very quick in coming back with a response. Apparently mpathadm has serveral associated bugs, and for accurate reporting, i'd be better off using:

Code:
# luxadm display /dev/rdsk/c6t60000970000298700009533031333037d0s2
DEVICE PROPERTIES for disk: /dev/rdsk/c6t60000970000298700009533031333037d0s2
  Vendor:               EMC
  Product ID:           SYMMETRIX
  Revision:             5874
  Serial Num:           700009307008
  Unformatted capacity: 98305.312 MBytes
  Read Cache:           Enabled
    Minimum prefetch:   0x0
    Maximum prefetch:   0xffff
  Device Type:          Disk device
  Path(s):

  /dev/rdsk/c6t60000970000298700009533031333037d0s2
  /devices/scsi_vhci/ssd@g60000970000298700009533031333037:c,raw
   Controller           /devices/pci@3,700000/SUNW,emlxs@0,1/fp@0,0
    Device Address              50000975f0002585,16
    Host controller port WWN    10000000c978d0c3
    Class                       primary
    State                       ONLINE
   Controller           /devices/pci@3,700000/SUNW,emlxs@0,1/fp@0,0
    Device Address              50000975f0002589,16
    Host controller port WWN    10000000c978d0c3
    Class                       primary
    State                       ONLINE
   Controller           /devices/pci@13,700000/SUNW,emlxs@0,1/fp@0,0
    Device Address              50000975f0002581,16
    Host controller port WWN    10000000c978d125
    Class                       primary
    State                       ONLINE
   Controller           /devices/pci@13,700000/SUNW,emlxs@0,1/fp@0,0
    Device Address              50000975f000258d,16
    Host controller port WWN    10000000c978d125
    Class                       primary
    State                       ONLINE

Which indeed showed all paths online as expected. To fix mpathadm, they've suggested I apply patch
143643-08, and a series of others. But after removing disk/configuring once more, it's reporting back ok.

Thanks for everyone's help. Smilie
Login or Register to Ask a Question

Previous Thread | Next Thread

10 More Discussions You Might Find Interesting

1. UNIX for Beginners Questions & Answers

Single LUN or multiple smaller LUNs for NFS sharing

We are using Solaris 11 Logical Domains. We NFS share a disk backup area from one LDom to other servers. The share is on a single 1.5 Tb LUN presented to the Control Domain, and as a virtual disk on the NFS server LDom. A 1.5 Tb zpool has been created on the LDom. We now need to provide more disk... (2 Replies)
Discussion started by: SallyB
2 Replies

2. AIX

Changing AIX LUN paths

Hello, Looking for some guidance to changing how an AIX 6.1 server sees its luns. Presently they're connected via 1GB iscsi but I want to change it to 4GB fiber channel. Anyone done this before and if so is it an easy step? I'm thinking this can be done in this order: unmount the... (5 Replies)
Discussion started by: bbbngowc
5 Replies

3. Solaris

Luxadm display not showing a LUN as expected

Hi all, I have the following LUN which is showing OK with its two paths: DEVICE PROPERTIES for disk: /dev/rdsk/c23t500507630A3BC579d123s2 Vendor: IBM Product ID: 2107900 Revision: 3010 Serial Num: 75WB0114900 Unformatted capacity:... (13 Replies)
Discussion started by: ludiegu
13 Replies

4. Solaris

How to avoid seeing various paths to just one LUN?

Hi I have a server running solaris 10 in which the NetAppadmin has just presented one LUN of 200Gb, but when running: bash-3.2# format Searching for disks...done c3t500A09828DE3E799d1: configured with capacity of 199.94GB c3t500A09829DE3E799d1: configured with capacity of 199.94GB ... (5 Replies)
Discussion started by: fretagi
5 Replies

5. Solaris

LUn' unmapped from storage still showing on Solaris 10

How can i remove LUN's from solaris 10 those are unmapped from SAN? from storage side i got confirmation that they removed lun. but in my solaris box still it's visible. i tried below. root@globalares2.example.com #luxadm -e offline /dev/rdsk/c2t600D02310007D16C42FF09E24B5B8255d0s2... (7 Replies)
Discussion started by: bentech4u
7 Replies

6. UNIX for Advanced & Expert Users

[Solved] mpathadm list lu no output

When I try to configure my HBA ports on Solaris 10; I cant see any output. bash-3.2# mpathadm list lu bash-3.2# I checked /kernel/drv/fp.conf file. ... mpxio-disable="no"; ... What shall I check more? what shall I do? ---------- Post updated at 10:18 AM ---------- Previous update was... (0 Replies)
Discussion started by: hubatuwang
0 Replies

7. Solaris

Adding new luns to mpxio config - showing up twice in format.

Hi, I have a sun 440, it already has LUN's configured with MPXIO. I just had more LUNS added, I did devfsadm and cfgadm -al, and they're showing up twice in "Format". :confused: There are no new fiber cables or HBA's involved, same physical setup, just adding LUNS. Aren't they... (0 Replies)
Discussion started by: BG_JrAdmin
0 Replies

8. Solaris

luns

how to create luns plzzzzzz (1 Reply)
Discussion started by: nsusheelgoud
1 Replies

9. Solaris

I have LUN ID, how to find disk relate to that LUN ID?

I have a list of LUN ID, my task is to find if disk has been added or not. How do I do that? I have been searching the forum and not able to find answer. thanks (4 Replies)
Discussion started by: uuontario
4 Replies

10. Filesystems, Disks and Memory

vxdisk list output showing extra LUNs

hi fren i am facing this problem i have sun 6140 storage 30 volumes bt my vxdisk list output is showing 37 disk which shd nt b there has neone faced this proble earlier vxdiks list output is like this SUN6140-2_0 auto - - online c4t201700A0B84853C0d2s2 SUN6140-2_1 auto - - online... (13 Replies)
Discussion started by: bond_ankur
13 Replies
Login or Register to Ask a Question