Unsolved
This post is more than 5 years old
212 Posts
0
6993
Installation issues with a VNXe3300
Hi
We have tried to install a VNXe3300 dual SP.
The unit was shipped to us from the factory in Cork last week.
We have followed the instructions in the configuration worksheet.
Racked the unit
Cabled with Power end Ethernet cables
Power on
And the unit seemed to work fine
Then we used the connection utillity to make a config file and saved that to a USB stick
Inserted the stick in SP A, and waited for the SP status LED to go out, and removed the USB Stick
This is not the first VNXe we install, and all the others have worked fine.
The situation is now that both SP status LED are blinking first blue then amber then blue and so on, at a frequency about once every second...¨
I cannot find this pattern of blinking in the VNXe3300 hardware information guide or in any release notes
This has been going on for about 12 hours.
This morning we have tried shutting down the unit, but after reboot it comes up with the blinking LED's again
Can anyone explain what this form of blinking LED's mean, and how to solve it.
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
1
July 20th, 2011 04:00
Hi Jim,
The SP is stated to be in Service Mode when the LED blinks between blue and amber.
If possible, you could connect the console cable to one of the SP and run the following commands
svc_rescue_state -c
and reboot the SP with : svc_shutdown -r
Now verify if the LED stays solid Blue. If it does, repeat the same steps on the second SP.
If the LED still remains blinking blue and amber, please log a Service request with EMC support for us to investigate.
Regards,
Amal.
emcBT
15 Posts
1
July 20th, 2011 06:00
To add to this -
Jim, if you're on the serial connection as the service account and able to run service scripts, you could also run 'svc_sysstatus' and attach the output here. That would indicate why each SP is in service mode, and if there are associated service code hints on either SP that would help EMC support get your system back up and running faster.
Jim_Hegner
212 Posts
0
July 21st, 2011 05:00
I have now resolved the issue, but had to have EMC support lending a helping hand.
Are there any reasons why a EMC partner cannot or should not use all the svc_ commands...
Because the only thing support did was run one more command than the ones we had already tried...including the ones mentioned in the 2 earlier posts.
Jim_Hegner
212 Posts
0
July 21st, 2011 06:00
Hi
Would these issues not be shown by the svc_diag -state= extended command.
emcBT
15 Posts
0
July 21st, 2011 06:00
Hi Jim,
Great question. The service scripts that ship on the VNXe are intended to be used in service-engaged situations where either an authorized support person is running the service scripts themselves (locally or via remote access) OR a customer is running the scripts under the direction of an authorized support person.
A subset of the service script functionality is intended for end customer use and those commands are also exposed via the Unisphere GUI on the Service page. You need the service account password to access that page and those commands, but once in you can see the state of both SPs (if in service mode, explains why and gives hint codes for what specifically is failing) and you can perform common service operations like Data Collection ("Collect Service Info"), reboot the SPs, enable SSH, and a handful of others.
The functionality that lies in the service scripts/tools which are not on the Service GUI page in Unisphere are accessible by anyone with the service password, but should not be used unless a support person is engaged as there is a good chance there could be unintended consequences.
emcBT
15 Posts
0
July 21st, 2011 06:00
Oh, and one other thing to mention - any time that one or all/both of the SPs in a VNXe are in service mode, the first thing to do should be to pull up the Unisphere GUI and look at the service page to see what is wrong. Simply clearing the rescue state and rebooting the SPs will often mask some serious problem and it's important to first know the mode/state of each SP, and if any service hint codes are registered on either SP. Those will pinpoint the problem for authorized service personel and speed up the time it takes to resolve the situation.
emcBT
15 Posts
0
July 21st, 2011 07:00
The "rescue reason" codes that put the SP into service mode would be shown, and the hint code would be shown as well... but the Unisphere GUI is able to do a translation of the rescue reason codes into a Problem statement and an Action statement that is intended to be user-actionable. Besides that, you're right - it's roughly the same set of information.