How to re-image a storage processor?

How to re-image a storage processor?













Re-image a storage processor (SP) to safely fix problems with the system software that could not be resolved by rebooting the SP. Re-imaging analyzes the system software for errors and attempts to correct them instantly. Re-image an SP from the Service System page.


Your configuration settings and stored files will not be changed.


Re-imaging an SP involves placing it in Service Mode. When an SP is in service mode it stops servicing I/O to hosts. All storage servers on the SP fail over to the other SP, if it is healthy. By default, when re-imaging has completed, the storage servers fail back to the SP. If the Failback Policy is disabled, all storage servers on the SP will not fail back automatically and will remain on a single SP. Performance can degrade significantly when all storage servers reside on a single SP.


To re-image an SP from Unisphere GUI:


Click Settings > Service System.
Enter the service password to access the Service System page.
Under System Components, expand Storage System.
Select the SP you want to re-image. If the SP is in Service Mode, as indicated by the Mode field, skip to Step 9.
Under Service Actions, select Enter Service Mode, to place the SP in Service Mode.
Click Execute service action.
In the confirmation dialog box, click OK.


Caution! Do not perform any actions in Unisphere until this operation has completed.
Wait at least 10 minutes while the SP enters Service Mode.
With the SP selected, look at the Mode field, which will display Service Mode to indicate that the SP is in Service Mode.
Consider these points:
    By default, the Service System page will refresh every 60 seconds to display the current status and mode of the SP.
    Once in Service Mode, the Fault LED on the SP flashes amber and blue.
Under Service Actions, select Re-image.
Click Execute service action to start the re-image process.
Wait at least 20 minutes while the system re-images the SP.
Select the SP you re-imaged and reboot it to return it to Normal Mode.


To re-image an SP from CLI:


Note: SP that needs to be reimaged requires to reboot couple of times during this process. All services will failover and run on peer SP if it is in healthy state. If connected through SSH session, you may lose connectivity while SP is rebooting.


1. Connect to VNXe using SSH session (PuTTy) or serial port. If connecting through SSH, make sure you are connected to the right SP which you want to reimage.


2. Put the SP into service mode using the commands below.


         svc_rescue_state -s
         svc_shutdown -r


3. After the SP reboots into service mode use the following command to reimage.


      svc_reimage -r


4. SP will reboot again and will be reimaged.


5. If SP is still in service mode, use the following commands to return to normal mode.


         svc_rescue_state -c
         svc_shutdown -r


Currently there is no option available to reimage the SP through Unisphere CLI.


If re-imaging the SP does not fix the problem, go to the EMC Online Support website for support options. For details about the support options, go to the Support by Product page.




EMC Support Solution Number: 14404