This post is more than 5 years old
14 Posts
0
25257
July 10th, 2013 11:00
CX3-20 backplane cables out of sequence - office move jumbled them up!
Hi
We've just moved offices and our 7 rack CX3-20 has been re-installed in the new building but the sequence of which cable goes in which fibre port for the 4Gb backplanes has been lost! Duh!
Currently the Windows 2008 server hosts can see the correct disk drives and read & write to them but whilst I can login to the SPs Navisphere just shows a 'fault' on each SP and does not show any LUNS, drives or disks.
So, silly question, but is there anyway to retroactively discover what fibre should plug into which disk array? The diagram we used for the removal obviously is wrong!
Thanks
David
No Events found!
DELL-Sam L
Moderator
•
7.6K Posts
1
July 10th, 2013 14:00
Hello David,
What you need to do is to login to Navisphere or usphere (depending on your flare version) & right click on the CX3-20. There will be a list of choices and you can choose Connectivity Status. That will bring up a new window that will list out all HBA’s connections to the CX3-20 & which ports that the connections are being made on. All connections that are green are connections that are active. All with a blue arrow are registered but not active & any that are Gray are not seen by the CX3-20.
If you are able to verify that all your connections are good then what you may need to do is to restart the management server on the CX3-20. How you do that is to go to the IPaddress/setup (example 192.168.1.1/setup). Once there you will need to enter your username & password. Once logged in then you can select Restart Management Server. NOTE: will need to do this to both SP’s at the same time to ensure that the management server gets restarted so recommend that you have 2 browser windows open so you can do both at the same time. Once you select to restart the server you will have to wait about 1 minute and then you can log back into the navishpere again & check to see if you have any faults. If still having issues then will need to pull some SPcollects and will need to review to see what the issue is that is causing the faults.
Please let us know if you have any other questions.
Dev Mgr
4 Operator
•
9.3K Posts
1
July 10th, 2013 18:00
The CX3-20 has 1 backend bus that supports up to 8 DAEs. This means that you don't have to worry about which enclosure was on which backend bus.
However, you do need consistent cabling. So from the SPs SPA goes to the A-side on enclosure 0, SPB goes to the B-side on this enclosure. The next enclosure daisy chains off of that and you keep doing that till the last disk enclosure. Don't cross over from A to B or use a different order for the cabling between SPA and SPB (e.g. SPA goes 0, 1, 2 and SPB goes 0, 2, 1).
If the LUNs are accessible, then it is indeed probably not a bad idea to restart the management server on both SPs (within seconds of eachother). This does not affect LUN connectivity; it only restarted Navisphere.
Dev Mgr
4 Operator
•
9.3K Posts
1
July 12th, 2013 20:00
DAE 0 has to be the first one on the chain, as this is where the OS of the storage processors is location (on 4 of the first 5 drives (all 5 are used by the overall flare, but the OS itself is mirrored on disk 0 and 2 for SPA and disks 1 and 3 for SPB)).
After that the order is supposed to be flexible (as long as you do not change enclosure IDs), but I'd recommend to use normal ordering; 0, 1, 2, etc.
davidmatthewson_2000c9
14 Posts
0
July 11th, 2013 16:00
Thanks for both of those suggestions. I tried running up Navisphere (Java version) on a directly connected laptop and, after logging in with my creds, it 'sees' the SP (either A or B) but shows 'F' (fault) against both the 'domain' and the array. RBC should, as you say, bring up a menu which it does. Select 'Connectivity' and the browser just locks up with no result. I get similar results on both SPs.
Thanks for the advice about the backplane cables - just to be clear, it doesn't matter which 'sequence' the DAEs are in? [so long as SPA goes to 'A' side and vv?? I may have the DAEs in the wrong 'stack' order but that would not seem, you say, to matter??
Indeed, the Windows host 'sees' all the drives at a Windows level and they are all R/W OK.
So any ideas how to get the SPs (connected via a Brocade 200e) to 'see' the LUNS etc for managment?
Many many thanks
David
davidmatthewson_2000c9
14 Posts
0
August 2nd, 2013 07:00
Hi
OK, we'll close this off now & thank everyone for all their help.
We've learnt that it doesn't matter what sequence the DAEs are connected up in so long as (i) you keep the 'in' and 'out' sockets correctly matched up & (ii) keep SPA and SPB together - that is, do not cross anything over from left to righ on the back of the DAEs.
We've also confirmed that it can take 30 mins or more fro the SPs to recover from a reset and this is 'normal' (!?)
It seems we have a dead SP but at least now can use Navisphere and manage the SAN. As it's 5 years old + I guess it's time to junk it and replace it.
Many thanks for all the advice.
David