Verifying RAID Group Disk Order
The examples above are from an output of running the get Raid Group command from Navisphere Command Line Interface.
Both RAID Groups are configured as Raid type 1_0.
In an earlier blog we discussed the importance of configuring RAID 1_0 by separating the Data disks and Mirrored Disks across multiple buses and enclosures on the back of the Clariion. This diagram is to show how you could verify if a RAID 1_0 Group is configured correctly or incorrectly.
The reason we are showing the output of the RAID Groups from the command line is this is the only place to truly see if the RAID Groups were configured properly.
The GUI will show the disks as the Clariion sees them in the order of the Bus and Enclosure, not the order you have placed the disks in the RAID Group.
11 comments:
sanguy,
Our CX3-20 only has a single bus. We have tried to split our r1_0 over 2 enclosures as the next best thing, but the output from the CLI doesn't match your example.
What output would I expect to see in a single-bus system with the raid group split across 2 enclosures, with data drives on enclosure 0 and mirror drives on enclosure 1? Is there some other way I can verify this?
Regards,
Carl B.
honestly, with the CX3-20, you are not going to gain anything in terms of performance. The only thing you are protecting against if is the Bus 0 Enclosure 1 has a problem, you still have Bus 0, Enclosure 0. Also, please tell me you are not using the Vault Disks (0_0_0 through 0_0_4) in the Raid Group. But, the output, if set up properly should look something like this.
Bus 0 Enclosure 0 Disk 5
Bus 0 Enclosure 1 Disk 5
Bus 0 Enclosure 0 Disk 6
Bus 0 Enclosure 1 Disk 6
...and so on...
the best way to make this raid group would be from the CLI...and it would look something like this...
naviseccli -h IP Address of SP createrg # 0_0_5 0_1_5 0_0_6 0_1_6 -raidtypye r1_0
sanguy,
That is what we were trying to accomplish, isolate Enclosure X problems from Enclosure Y. No, we are not using the Vault Disks in the Raid Group. getrg lists all the Enclosure 0 drives first, then lists all the Enclosure 1 drives. Based on the example, and if it applies in this case, we didn't get the split we wanted of all data drives on Enclosure 0 and mirror drives on Enclosure 1. I'll try your suggestion and post an update. Thanks for a great blog.
Regards,
Carl B.
Sanguy,
Thanks for this post. I couldn't for the life of me remember how to check this from the class 2 years ago :) and I needed to create a mirror of stripes across two buses. Hope all is well with you and your family.
Scott M.
Sacramento
Sanguy,
Having read your blog yesterday I thought I'd look at a couple of our raid 1_0 RGs and sure enough a getrg shows disks all over the place.
So I create a new RG through Navisphere and add my disks one at a time. Bus3, Bus2, Bus3, Bus2 etc so I had my data/mirror luns correctly spread accross two shelves. I run the getrg and it looks great. I add the LUNs and suddenly it looks like this:
RaidGroup ID: 62
RaidGroup Type: r1_0
RaidGroup State: Explicit_Remove
Valid_luns
List of disks: Bus 3 Enclosure 4 Disk 2
Bus 3 Enclosure 4 Disk 3
Bus 3 Enclosure 4 Disk 4
Bus 2 Enclosure 4 Disk 2
Bus 2 Enclosure 4 Disk 3
Bus 2 Enclosure 4 Disk 4
List of luns: 607 608
Can you shed any light? If I remove the LUNs it goes back to:
RaidGroup ID: 62
RaidGroup Type: Unbound
RaidGroup State: Explicit_Remove
Valid_luns
List of disks: Bus 3 Enclosure 4 Disk 2
Bus 2 Enclosure 4 Disk 2
Bus 3 Enclosure 4 Disk 3
Bus 2 Enclosure 4 Disk 3
Bus 3 Enclosure 4 Disk 4
Bus 2 Enclosure 4 Disk 4
List of luns: Not Available
NST Version : 6.28.53.2.3
Subsystem Serial Number : CK20008******
Subsystem Model Number : CX3-80
FLARE Operating Environment Rev : 03.26.080.5.016
Enclosure type : DAE2-ATA
Enclosure Speed : 2Gb/s
Hope you can help clear this up.
Rgds
D.
Hi sanguy and Anonymous,
Care must be taken when interpreting the output for the
naviseccli -h xxx.xxx.xxx.xxx getrg nn –disk
command when listing a R1_0 Raid Group.
This is because the output is different depending on if a LUN has been bound in the Raid Group or not.
Assuming you issue the command
Naviseccli –h xxx.xxx.xxx.xxx \
createrg 8 \
0_3_0 1_4_4 \
0_3_1 1_4_5 \
0_3_2 1_4_6 \
-rm no \
-raidtype r1_0
Then
0_3_0 is primary1
1_4_4 is mirror1
0_3_1 is primary2
1_4_5 is mirror2
0_3_2 is primary3
1_4_6 is mirror3
If immediately after the Raid Group creation you issue the
naviseccli -h xxx.xxx.xxx.xxx getrg 8 –disks
command it will list the disks thus
List of disks:
Bus 0 Enclosure 3 Disk 0
Bus 1 Enclosure 4 Disk 4
Bus 0 Enclosure 3 Disk 1
Bus 1 Enclosure 4 Disk 5
Bus 0 Enclosure 3 Disk 2
Bus 1 Enclosure 4 Disk 6
So before the first LUN is bound the RAID structure does not actually exist and the disks are listed:
primary1
mirror1
primary2
mirror2
primary3
mirror3
However, when the first LUN is bound, the Raid structure is created on the Raid Group and then the query lists the disks thus:
List of disks:
Bus 0 Enclosure 3 Disk 0
Bus 0 Enclosure 3 Disk 1
Bus 0 Enclosure 3 Disk 2
Bus 1 Enclosure 4 Disk 4
Bus 1 Enclosure 4 Disk 5
Bus 1 Enclosure 4 Disk 6
So while LUNs are bound in the R1_0 Raid Group the disks are listed:
primary1
primary2
primary3
mirror1
mirror2
mirror3
The functions of the disks has not changed.
It is just that while a LUN is bound in the Raid Group all the primary disks are listed first and then all the mirrors.
Regard,
Frank
wow - great posts all, thanks much!
by any chance do you know where I might be able to pickup a replica watch? ;-)
I saw the same content here:-
http://www.emcclariion.com/training.html
yeah...not sure who or what emcclariion.com is
the getrg command will not list the disks in order anymore.
Bus 3 Enclosure 0 Disk 0
Bus 3 Enclosure 0 Disk 1
Bus 3 Enclosure 0 Disk 2
Bus 3 Enclosure 0 Disk 3
Bus 0 Enclosure 1 Disk 0
Bus 0 Enclosure 1 Disk 1
Bus 0 Enclosure 1 Disk 2
Bus 0 Enclosure 1 Disk 3
are you sure EMC has fixed this? I don't think so.
One way check for the mirrored disks is check the size. The data and mirror disk must have the same size.
Post a Comment