chjatwork

updated

10 years ago

C

chjatwork

2 Intern

356 Posts

0

28893

November 17th, 2016 06:00

Internal Boot Drive replacement

Community,

Any chance anyone would have the written procedures for how to replace the boot disk?  I believe the nodes are 108NL.  I know its a simple replacement but just want to make sure I have the written guidelines for reference and to back me up with my process.

I am getting these errors and know that I need to replace this boot disk. 

24 Drive at Internal J3/ad4 wear_life threshold exceeded: 100

Thank you,

  • sjones51

    252 Posts

    20433

    2

    Posted November 17th, 2016 07:00

    Hi chjatwork,

    Is this cluster still covered under an Isilon Support Contract? If so, the replacement should be done by a certified engineer as boot drives are not considered to be a customer replaceable unit. You can open a service request here for the replacement if one has not been automatically created:

    https://support.emc.com/SRcreate

    Otherwise there is documentation located here:

    https://support.emc.com/docu44208

    Be sure to verify that the node is indeed a 108NL. A mistake made in replacing the boot drive could lead to data loss.

  • chjatwork

    2 Intern

    356 Posts

    20434

    0

    Posted November 18th, 2016 04:00

    sjones5,

    Just in case its the NL400 chassis would you happen to have that replacement document too?  I totally understand, but this customer I am working for has experienced numerous boot disk failures on this cluster and is accustom to just doing this themselves.  So I am trying to make sure I have the information I need to continue this process.  Let me know about the document for the NL400.

    Many thanks!

  • sjones51

    252 Posts

    20434

    1

    Posted November 21st, 2016 08:00

    Hi chjatwork,

    While I am sure the customer is completely capable, if they have a support contract, they are not to replace the boot drives on their own. If they do, they risk voiding that support contract. The standard disk drives are customer replaceable, but the boot drives are not. Anything that requires opening up the node, must be done by a certified engineer.

    Here is the NL400 documentation as well:

    https://support.emc.com/docu44286

  • chjatwork

    2 Intern

    356 Posts

    20433

    0

    Posted November 22nd, 2016 07:00

    sjones5,

    They are fully aware of this that's why they have me on staff as a resident.  I am able to replace these disk.  I just would like to have the documentation on hand for reference and best practice.  I am glad that my account level can access the links so all is well.

    Many thanks again for following up with me!

  • jjacobu

    5 Posts

    20487

    0

    Posted November 23rd, 2016 04:00

    We had local EMC Support replace our boot drive but the support engineer generated replacement documentation from SolVE desktop application.

  • jarec007

    2 Posts

    20487

    0

    Posted August 3rd, 2017 10:00

    I was wondering if you have this documentation for the x400?  I am also looking for the part number for the boot drives if you have that as well.

    Thanks

    The service has elapsed on this cluster

  • jjacobu

    5 Posts

    20487

    0

    Posted August 4th, 2017 06:00

    I found this part number in my notes: Drive Part Number: 403-0114-01


    isi_for_array -s isi_radish -a /dev/ad* | grep -E "Percent Life" | grep -v Used | awk '{print $1 $2 $3 $4 sprintf( "%d","0x"$9 )}'

    Will display wear life for each drive in every node in the cluster, which is nice

  • sjones51

    252 Posts

    13324

    1

    Posted August 4th, 2017 08:00

    Hi jarec007,

    The documentation for the X400 is here: https://support.emc.com/docu44413

    As far as finding a replacement part, you have a couple options. You can do a Time and Materials quote for the part replacement through Isilon to have them replace it. They will need the serial number of the node in question.


    If you are looking to purchase a boot drive from a third party, I don't know that the Isilon part number would be valid outside of Isilon's dispatch system.You could however find a compatible model number. If you know what firmware version you are on, check the release notes. Otherwise you should be able to see what model you currently have with this command:

    #isi_radish -q /dev/ad[0,1,2,3,4,7]

    As a side note, if you are running anything below 8.0, you can use [2,3,4,7]. In 8.0 OneFS started streamlining the boot drive names to ad0 and ad1. The command above should work on your X400 though. It doesn't change anything, It just provides an informational output.

    Just know that if you do go the third party route, Isilon won't assume any liability for it.

  • ebaxter1

    2 Posts

    20487

    0

    Posted September 28th, 2017 05:00

    isi_for_array -s isi_radish -a /dev/ad* | grep -E "Percent Life" | grep -v Used | awk '{print $1 $2 $3 $4 sprintf( "%d","0x"$9 )}'

    Hello trying to run this command on a cluster running Isilon OneFS v7.1.1.4. the output of isi_radish -a /dev/ad* does not contain "Percent Life"  I do see this line...

    Percent of Total Wear: 01/64 vs 00 (raw 000000000000) [flags 0003]

    how can I tell what percentage of life is left on a cluster that is running OneFS v7.1.1.4?

  • jjacobu

    5 Posts

    10928

    0

    Posted September 29th, 2017 06:00

    You'll probably want to confirm with support but I believe (raw 000000000000) is the hexidecimal value for the status of life remaining, and in your case it is zero.