Linux MPIO - CX300 Missing Path

possible zoning issue? Have you checked on each CX300 in the Connectivity Status screen to see if all the paths are logged in and registered? Also, if you're using MPIO on Linux what is the failover mode setting for the LUNs on the array?'

See Knowledgebase article emc99467

Parameter PowerPath DMP DM-MPIO
Initiator Type 3 (CLARiiON Open) 3 (CLARiiON Open) 3 (CLARiiON Open)
Arraycommpath 1 (Enabled) 1 (Enabled) 1 (Enabled)
Failovermode 1 ¿or- 4 note 11 1 ¿or- 2 note 9 1 ¿or- 4 note 10
UnitSerialNumber Array Array Array

9.

For Linux installations using VERITAS VxVM, set failovermode as follows:
¿ VERITAS VxVM 4.x or 5.x, with PowerPath: failovermode MUST be set as
required by PowerPath
¿ VERITAS VxVM 4.x without PowerPath: failovermode MUST be set to 2
¿ VERITAS VxVM 5.x without PowerPath: failovermode MUST be set to 1

10.

For Linux DM-MPIO, failovermode settings are as follows:
¿ FLARE 24 or earlier: failovermode MUST be set to 1
¿ FLARE 26 or later:− Novell SLES:
-Novell SLES 10.1 (SP1) or above: failovermode may be set to 1
for active/passive behavior, or
-Novell SLES 10.1 (SP1) or above: failovermode may be set to 4
for ALUA behavior
-Novell SLES 10.0 (no SP) or below: failovermode MUST be set
to 1

− RedHat RHEL 4.x:

-RHEL 4.6 or above: failovermode may be set to 1 for
active/passive behavior, or
-RHEL 4.6 or above: failovermode may be set to 4 for ALUA
behavior
-RHEL 4.5 or below: failovermode MUST be set to 1

− RedHat RHEL 5.x:

-RHEL 5.1 or above: failovermode may be set to 1 for
active/passive behavior, or
--RHEL 5.1 or above: failovermode may be set to 4 for ALUA
behavior
-RHEL 5.0: failovermode MUST be set to 1
− Asianux:
-Asianux 3.1 or above: failovermode may be set to 1 for
active/passive behavior, or
-Asianux 3.1 or above: failovermode may be set to 4 for ALUA
behavior
-Asianux 3.0: failovermode MUST be set to 1

− Other Linux distributions: failovermode MUST be set to 1

11.

For PowerPath for Solaris, Linux and Windows, failovermode settings are as follows:
¿ FLARE 24 or earlier: failovermode MUST be set to 1
¿ FLARE 26 or later:
- PowerPath 5.0 or earlier: failovermode MUST be set to 1
- PowerPath 5.1 or later: failovermode may be set to 4 for ALUA behavior, or
- PowerPath 5.1 or later: failovermode may be set to 1 for active/passive
behavior

glen

To add to what Glen mentioned, right click the array and click connectivity status. In here you can sort different columns, but basically you want to make sure all of your paths are logged in and registered. If your missing paths, most likely zoning. If things are logged in and not registered, then you need to register them. If they are logged in and registered, and you still don't have all paths to the lun, then most likely you don't have all paths in the storage group.

Under storage groups expand the group in question and expand the host section. In here right click the host and click connectivity status. Are there paths in ~management? If so then they don't have access to the luns. To fix this just hit reconnect and do your magic on the host side to get all paths to be recognized.

Usually the ~management thing happens when you zone and register a path and put it to the storage group, then finish the zoning and registering. At this point the Clariion does not assume you want those paths to have access to the luns, so it doesn't grant access.

Another way you can get into this situation is if you replace an HBA on the host. You'll need to zone, register, and add the paths to the storage group.

-Ryan

Just wanted to say thanks to both for your responses, so much has been going on at this site I can only troubleshoot this problem about once a week (fortunately I haven't had a failure of the primary path). I'm using your suggestions to try and track down the missing path.