This post is more than 5 years old
1.3K Posts
0
1974
number of available mset mirrors
"The operation cannot be performed because the number of available mset mirrors would be exceeded"
i am trying to understand more on the limitation of the number of avaliable msets..
i am trying to understand more on the limitation of the number of avaliable msets..
xe2sdc
2.8K Posts
0
October 6th, 2009 07:00
With SYMAPI_SYSCALL_RANGES_HEURISTIC you can enable "range syscalls" that will speed up command execution. But if you have a binfile with holes (non contiguous symdevs) syscalls may fail with strange and exotic errors. That's why you can always disable heuristic ranges and use good old "lists" of devices. Slower, but reliable
MarcT2
131 Posts
0
January 8th, 2008 09:00
Scratch the bit about "symdev show" - it doesn't appear to show the presence of an established BCV in the mirror set information.
MarcT2
131 Posts
1
January 8th, 2008 09:00
Each Symmetrix device has 4 "mirror positions". A standard mirrored device uses two of these (one for each copy).
SRDF uses an additional mirror position (on both the R1 & R2), so our device is now up to 3. Establish a BCV and all 4 mirror positions are used. This means a second concurrent BCV or RDF volume cannot be established, since there's nowhere for it to go.
A traditional hotspare also takes up a mirror position whilst invoked, so a mirrored RDF volume with a spare invoked cannot then have a BCV established since there is no free mirror position.
You can use "symdev show " to see what's in the mirror positions, you'll get something like:
Mirror Set Type : [Data,Data,N/A,N/A]
in the output.
Hope that's useful!
Regrads,
Marc
SKT2
1.3K Posts
0
January 8th, 2008 10:00
This could be mostly due to a hot spare being invoked at that point in time since we never go beyond one SRDF/BCV copy.
But i need to confirm the presnce of a hot spare if that is the case.
josevtome
22 Posts
0
August 20th, 2009 01:00
If I have a volume in a failed disk, and this volume have already it's 4 mirror possitions occupied by SRDF, TF/Mirror operations, etc, can there be any problem while invoking the Hotspare operation?
Lately the Hotspare scripts are taking very long in my platform, 12+ hours to finish sync. Is this normal, or could the lack of mirror possitions be affecting this operations?
SKT2
1.3K Posts
0
August 30th, 2009 01:00
ms_spare_mir4, ms_allow_dynamic_mir and ms _smmf_std flags .
what is the size of the failed disk and what is the code version of the array?
MikeMac1
292 Posts
1
September 1st, 2009 10:00
have already it's 4 mirror possitions occupied by
SRDF, TF/Mirror operations, etc, can there be any
problem while invoking the Hotspare operation?
Symm devices only have 4 available mirror positions so if all 4 mirror positions are currently in use a hot spare will not invoke.
my platform, 12+ hours to finish sync. Is this
normal, or could the lack of mirror possitions be
affecting this operations?
Reads and writes from the host take priority over a hot spare syncing. There are many variables as to why a hot spare could be taking a longer time than suspected to sync up. You mention TF/Mirror, are you also currently doing BCV operations at the same time that the hot spare is syncing?
mvs3
1 Message
0
September 9th, 2009 02:00
- A way to see / visualise how many mirror positions are in use.
- Clarification if TimeFinder Clone volumes occupy one of the 4 mirror positions (or not).
- Clarification if free mirror position(s) are needed to use SymOptimizer
xe2sdc
2.8K Posts
0
September 9th, 2009 03:00
If you have a few spare minutes don't forget introducing yourself in coffee break area..
1) a way to see/visualize how many mirror position are in use.
Just issue "symdev -sid XXX show 0BAD" and look carefully at the output. Unfortunatly I can't show you an example output but I guess you can easily spot the required details.
Or simply just post output of above command (assuming you want to know everything about your BAD device) and we'll help you in finding the details you need.
2) Clone won't ever use mirror positions. Fullstop.
3) I guess MikeMac will happily answer with plenty of details
RobertDudley
448 Posts
0
September 9th, 2009 05:00
xe2sdc
2.8K Posts
0
September 9th, 2009 05:00
MikeMac1
292 Posts
0
September 9th, 2009 06:00
pr12282
14 Posts
1
September 10th, 2009 02:00
An alternate suggestion to address your issue, are :
1. It is possible to stop hotspare being invoked in the event of disk failure. Then plan to replace the disk when there is no impact to backup. However, this has increased risk of data loss in the event of the surviving disk fails while you wait for the backup to complete for replacement of the already failed disk
2. Use TF/Clone instead of TF/Mirror,which does not occupy any mirror position
STD+R1+BCV uses 2+1+1=4 mirror positions
-Raju
josevtome
22 Posts
0
September 10th, 2009 02:00
xe2sdc
2.8K Posts
0
September 10th, 2009 02:00
It's quite strange that the sync freezes 'till drive is replaced. Are you running in ACP mode?