This post is more than 5 years old
4 Posts
0
2050
August 11th, 2016 11:00
PPME error(22): while migrating MSCS devices
Good Day All,
I am in the process of migrating our SQL clusters from VMAX to XtremIO. I've created the XIO devices, added them to the SQL host and completed the steps required to start the sync. I have successfully gotten a few of the devices to the "SourceSelected" state, however as I wait for the remaining devices to sync, the some of the ones that were already in the "SourceSelected" state end up faulting and I get this message - c7 harddisk10 harddisk18 HostCopy(cl) sourceSelected targetLUfault PPME error(22): Target faulted, must abort migration. I then have to abort and then start the sync all over again. How do I deal with this issue? Is it okay once one of the devices is in a "SourceSelected" state to go ahead and commit that device? If I do that then some of the cluster devices will be on XIO while some will still be on VMAX until the sync has completed and I am able to commit. Any advice would be appreciated. I have not contacted support on this as of yet, figured I'd reach out to the community here for advice first. Thanks in advance.
Looney
In the first query below you can see that "c7" is in SourceSelected state. The second query was run several hours later just to check on the status of the syncing devices. As you can see "c7" is now faulted. I aborted and restarted the sync, then set the throttle to 4 (which is 13%).
C:\Users\adm-blooney>powermig -query info -all
===========================================================
Hnd Source Target Tech State
===== ========== ========== ============ ==============
c1 harddisk9 harddisk17 HostCopy(cl) sourceSelected
c3 harddisk11 harddisk19 HostCopy(cl) syncing(2%)
c4 harddisk5 harddisk22 HostCopy(cl) sourceSelected
c5 harddisk7 harddisk20 HostCopy(cl) syncing(6%)
c6 harddisk15 harddisk21 HostCopy(cl) syncing(7%)
c7 harddisk10 harddisk18 HostCopy(cl) sourceSelected
C:\Users\adm-blooney>powermig -query info -all
==========================================================================
Hnd Source Target Tech State DevFault
===== ========== ========== ============ ============== =============
c1 harddisk9 harddisk17 HostCopy(cl) sourceSelected
c3 harddisk11 harddisk19 HostCopy(cl) syncing(7%)
c4 harddisk5 harddisk22 HostCopy(cl) sourceSelected
c5 harddisk7 harddisk20 HostCopy(cl) syncing(20%)
c6 harddisk15 harddisk21 HostCopy(cl) syncing(24%)
c7 harddisk10 harddisk18 HostCopy(cl) sourceSelected targetLUfault
PPME error(22): Target faulted, must abort migration
C:\Users\adm-blooney>powermig abort -handle c7
Abort migration for handle c7 ? [yes]/no: y
C:\Users\adm-blooney>powermig sync -handle c7
Start sync for handle c7 ? [yes]/no: y
C:\Users\adm-blooney>powermig throttle -handle c7 -throttlevalue 4
Set throttle option(s) for handle c7 ? [yes]/no: y
C:\Users\adm-blooney>powermig -query info -all
===========================================================
Hnd Source Target Tech State
===== ========== ========== ============ ==============
c1 harddisk9 harddisk17 HostCopy(cl) sourceSelected
c3 harddisk11 harddisk19 HostCopy(cl) syncing(7%)
c4 harddisk5 harddisk22 HostCopy(cl) sourceSelected
c5 harddisk7 harddisk20 HostCopy(cl) syncing(21%)
c6 harddisk15 harddisk21 HostCopy(cl) syncing(25%)
c7 harddisk10 harddisk18 HostCopy(cl) syncing(0%)
Mich
86 Posts
0
August 11th, 2016 18:00
Hi Looney, in normal circumstances yes you could commit with some devices still syncing etc.
Considering you had a TargetLU fault, that needs to be investigated first.
TargetLU would have to be one of the LUNs for the XtremIO.
I would be apprehensive to complete any migration until you understand why the target faulted.
Open a Service Request and ask support to review the EMCReports once collected.
Also make sure you have read the XtremIO Host Configuration Guide as well. Search for "docu56210".
Regards, Michel.
blooneytoonz
4 Posts
0
September 20th, 2016 13:00
Good day Mitch,
So sorry for the late reply. I did indeed open an SR. I supplied them with host grabs and they confirmed your suspicions regarding a fault on the target luns. It was nothing to be alarmed about, just an I/O error. I restarted the ones that faulted and monitored until each completed. Once each completed I committed them and life is good. Thanks for the advice.
Looney
Mich
86 Posts
0
September 20th, 2016 14:00
Great to hear Looney, thanks for the follow up.
Regards, Mich.