Start a Conversation

This post is more than 5 years old

Solved!

Go to Solution

2980

June 21st, 2012 14:00

Static vs. dynamic persistency / assignment (FCID)

Can anyone clearly explain the difference between static and dynamic persistency / assignment when creating persistent fcids on mds?

I'm not asking about persistent fcids or static domains or enabling persistent fcids on a vsan ... I'm talking about an already persistent fcid that has an assignment of "dynamic" vs. one that has an assignment of "static" ... or creating a new persistent fcid and selecting "dynamic" or "static" as the assignment type. What's the difference?

June 22nd, 2012 01:00

Hello moe2197,

This determines how persistent FCIDs that are not in use behave with the purge command. Dynamic entries are deleted, static entries are not deleted. There is a section in the MDS CLI guide titled "About Persistent FCID Selective Purging" that may help. The aim will be to purge stale entries, in order to obtain more free entries. This use case is discussed in another document (not the CLI guide) - "System Messages and Recovery Pocedures for the Cisco MDS 9000 Family" - the system/error message in question being "FCDOMAIN-5-FCID_PERSISTENT_TBL_HALF_FULL".

Best regards, Richard.

23 Posts

July 5th, 2012 12:00

Thanks, Richard!

1 Rookie

 • 

5.7K Posts

July 6th, 2012 03:00

Now suppose I have an HP-UX host which benefits greatly from persistent FCIDs and I need to replace the switches. Is it ok to give the new siwtch the same domain id, set it to fixed and manually set the FCID for the wwn of the HP-UX HBA that will be connected to this one? Or will HP-UX still somehow see that something has changed with concequences to LUN connectivity?

July 6th, 2012 09:00

Hello TrippleR,

Yes, if the FCID is managed correctly then HPUX will be happy. So, as you reference, set the fcdomain (logical switch within the VSAN) domain ID manually (static, not preferred), and manually configure the 24-bit FCID (the command will fail if the first byte of the 24-bit address does not match the domain ID set in the first step).

I know you know all the steps, but for the general benefit of the forum, the steps are detailed in "EMC RecoverPoint - Deploying RecoverPoint with SANTap and SAN-OS" (search on fcdomain or references to HP-UX and AIX).

Having said all this, later releases of HP-UX should cope with 24-bit FCID changes due to agile addressing technology (similar to dynamic tracking in AIX). Note that I have not tested this, but this is my understanding from documentation. Solaris was probably the leader here, with the WWN used as the SCSI target in the logical device path - very nice indeed (/dev/dsk/c#t d#).

Cheers, Richard.

1 Rookie

 • 

5.7K Posts

July 9th, 2012 01:00

Thank you for the good explanation, Richard

No Events found!

Top