Yes, I do. We were able to get the lun IDs back-- that's not the issue. I'm just wondering how the system would have lost these maps to begin with, and where it might store them (as a clue to how it lost them).
________________________________ From: "Fright, Ian" Ian.Fright@netapp.com To: Fred Grieco fredgrieco@yahoo.com; "Klise, Steve" klises@sutterhealth.org; "Learmonth, Peter" Peter.Learmonth@netapp.com; steve klise sklise@hotmail.com; Raj Patel phigmov@gmail.com; Andrew Laurence atlauren@me.com Cc: "toasters@teaparty.net" toasters@teaparty.net Sent: Friday, November 16, 2012 10:19 AM Subject: RE: LUNs lost initiator setting after a power outage
Do you have a recent ASUP? :> lun show –m lun show [ -v | -m | -c ] [ all | mapped | offline | online | unmapped | staging | -g initiator_group | -n node | lun_path ] Displays the status (lun_path, size, online/offline state, shared state) of the given lun or class of luns. With the -v flag supplied, additional information (comment string, serial number, LUN mapping) is also displayed. With the -m flag supplied, information about lun_path to initiator_group mappings is displayed in a table format. With the -c flag supplied, information about lun cloning status is displayed. A specific lun can be indicated by supplying its lun_path. When an initiator_group is specified, status is reported for all luns that are mapped to the initiator group. When a node is specified, status is reported for all luns that are mapped to initiator groups which contain that node. When staging is specified, information about the temporary luns preserved in the staging area is reported. Mapped luns are ones with at least one map definition. A lun is online if it has not been explicitly made offline using the lun offline command. -Ian From:toasters-bounces@teaparty.net [mailto:toasters-bounces@teaparty.net] On Behalf Of Fred Grieco Sent: Friday, November 16, 2012 8:58 AM To: Klise, Steve; Learmonth, Peter; steve klise; Raj Patel; Andrew Laurence Cc: toasters@teaparty.net Subject: Re: LUNs lost initiator setting after a power outage Just had this happen after a power failure. IBM N6040 (V3140) with Ontap 7.3.3. Most lun mappings were lost, but the igroups remained. The system was up and down several times over a period of several hours, with some of the aggregates "missing."
1) Where does the netapp actually store the lun mappings?
2) Any idea why this would happen?
We name our luns with lun # and host name, and have less than hundreds of Luns, so manually remapping was not impossible...
TIA,
Fred
________________________________
From:"Klise, Steve" klises@sutterhealth.org To: "'Learmonth, Peter'" Peter.Learmonth@netapp.com; steve klise sklise@hotmail.com; Raj Patel phigmov@gmail.com; Andrew Laurence atlauren@me.com Cc: "toasters@teaparty.net" toasters@teaparty.net Sent: Tuesday, October 9, 2012 2:29 PM Subject: RE: LUNs lost initiator setting after a power outage I would agree with Peter. PowerShell is the ticket. This is the one liner I have setup. This assumes you have a network share called server with a network share of netapp_luns.. It creates a text file with the server name, dumps to the local C:, then copies to a network share.. @echo off sdcli disk list >"c:%computername%_disks.txt" copy "c:%computername%_disks.txt" \server\netapp_luns From someone who told me once.. hint Peter.. “Share and enjoy!” From:toasters-bounces@teaparty.net [mailto:toasters-bounces@teaparty.net] On Behalf Of Learmonth, Peter Sent: Tuesday, October 09, 2012 4:48 AM To: steve klise; Raj Patel; Andrew Laurence Cc: toasters@teaparty.net Subject: RE: LUNs lost initiator setting after a power outage Igroups and LUN mappings are captured in autosupport. You might consider adding a data capture email that receives things like autosupports and having your NetApp devices copy asup there as well as to NetApp. Even if you don’t capture it yourself, anybody at NetApp including support and your account team should be able to look that up and send it back to you. Another thought is PowerShell toolkit has cmdlets (Get-NaLun, Get-NaLunMap, Get-NaIgroup) for capturing this data. You could probably even save it as an object or XML and use that to audit changes or even script fixes for when weird stuff happens. For those unfamiliar with the NetApp PowerShell Toolkit, check out the following: https://communities.netapp.com/community/products_and_solutions/microsoft/po... https://communities.netapp.com/docs/DOC-8285 https://communities.netapp.com/community/products_and_solutions/microsoft/po... I hope this helps! Peter From:toasters-bounces@teaparty.net [mailto:toasters-bounces@teaparty.net] On Behalf Of steve klise Sent: Tuesday, October 09, 2012 6:17 AM To: Raj Patel; Andrew Laurence Cc: toasters@teaparty.net Subject: RE: LUNs lost initiator setting after a power outage For what its worth, I have a scheduled AT job on the windows box that dumps the snapdrive connected LUN info to the local C:\ as well as a network share. That way if I loose a server, I know what was mapped, drive letters (if applicable).. Its my self documentation for LUNS on my windows boxes. Real easy to setup..
________________________________
Date: Tue, 9 Oct 2012 17:02:10 +1300 Subject: Re: LUNs lost initiator setting after a power outage From: phigmov@gmail.com To: atlauren@me.com CC: toasters@teaparty.net
Sounds similar to my experience a few years ago after a controller upgrade. http://www.gossamer-threads.com/lists/netapp/toasters/10436#10436 The initiators disappeared and I spent a decent chunk of a weekend putting them back. Lucky I had screenshots and terminal dumps as I didn't have backups of this info. Cheers, Raj. On Tue, Oct 9, 2012 at 2:54 PM, Andrew Laurence atlauren@me.com wrote: So, this was fun.
After a power outage on Saturday, our Exchange 2010 mailbox servers (all hosted on VMWare ESX) came up with only some of their LUNs. These LUNs are hosted on our HA-paired FAS2040 (OnTAP 7.3.4); each LUN hosts an Exchange database. Try as we might, we couldn't get these LUNs to mount.
After opening a support case with each of NetApp and Microsoft, and conferencing in both engineers, and several hours, the Microsoft engineer finally suggested we check the target LUNs on the filer. Turns out that each of the missing LUNs no longer had the setting for the mailbox servers' initiators. After checking that box the LUNs mounted without issue.
Anyone seen this happen before? The NetApp engineer didn't have any immediate ideas, and I haven't had luck finding a match in the list archives or in NOW.
Thanks, Andrew
-- Andrew Laurence Office of Information Technology atlauren@uci.edu University of California, Irvine
_______________________________________________ Toasters mailing list Toasters@teaparty.net http://www.teaparty.net/mailman/listinfo/toasters
_______________________________________________ Toasters mailing list Toasters@teaparty.net http://www.teaparty.net/mailman/listinfo/toasters
_______________________________________________ Toasters mailing list Toasters@teaparty.net http://www.teaparty.net/mailman/listinfo/toasters