Servicemag resume failed

[PCD]

Do you know a way to terminate a servicemag start  command?

[Torsten.]

It is degraded now, not failed.

How about

cli% showpd -c -p -cg 0

[PCD]

Here is the info:

CLD1-3PAR-InServ-01 cli% showpd -c -p -cg 0
         ---- Normal Chunklets ----- ---- Spare Chunklets ----
         -- Used -- ---- Unused ---- - Used - ---- Unused ----
ID Total    OK Fail Free Uninit Fail  OK Fail Free Uninit Fail
 0  1488  1018    0  377      0    0   0    0   93      0    0
 1  1488  1020    0  375      0    0   0    0   93      0    0
 2  1488  1020    0  375      0    0   0    0   93      0    0
 3  1488  1019    0  376      0    0   0    0   93      0    0
 4  1488  1018    0  377      0    0   0    0   93      0    0
 5  1488  1019    0  376      0    0   0    0   93      0    0
 6  1488  1019    0  376      0    0   0    0   93      0    0
 7  1488  1019    0  376      0    0   0    0   93      0    0
34  3724     0    0  225   3216    3   0    0  280      0    0
35  3724  3439    0    5      0    0 280    0    0      0    0
39  3724  3430    0   14      0    0  10    0  270      0    0
44  3724  3439    0    5      0    0 280    0    0      0    0
--------------------------------------------------------------
12 26800 18460    0 3257   3216    3 570    0 1294      0    0
CLD1-3PAR-InServ-01 cli%

[Dennis Handly]

>we want to cancel this process  on cage 0 and magazine 9 and restart the disk replacement process again. And would like your advise if this command:

> servicemag resume unmark -cg 0 -mg 9,  would do the job?

You would probably need to talk to Support about that.

And as long as "servicemag resume" or showpd can't talk to the disk, there isn't much you can "restart".

Note: unmark and resume are mutually exclusive.

What does this show:  showpd -state 34

>34  0:9:0    953344   3724   225   280 2210000A33008420 failed 0:0:1- 1:0:1-   N

You may need to try to manually onloop the mag, since the "-" above says the ports are disabled:

controlmag onloop cage0 9

If that doesn't work, you may have to manually reset the two ports.

Note:  There are 12 disks on these two ports!  So you should probably wait between the two commands.

controlport rst 0:0:1

controlport rst 1:0:1

[John Bowe]

How soon after inserting the magazine did you try to bring it online?  There needs to be a few minutes before initiating the "resume" command for all the drives to spin up.  If you initiate the command too soon, this error will occur.

[moinf]

We had the same issue on 1 13 in our 3par HP_3PAR 7400

It was solved when we triggered: 

servicemag resume -partial 1 13
........
Estimated time for relocation completion based on 29 seconds per chunklet is: 1 days, 14 minutes and 21 seconds
The cumulative output so far is:
servicemag resume -partial 1 13
.........
 checking for valid disks...
... disks in mag : 1 13
... normal disks: WWN [5000CCA0321642D7] Id [145] diskpos [0]
... not normal disks: WWN [5000CCA02D0E96E7] Id [87]
... verifying spare space for disks 145 and 145
... playback chunklets from pd WWN [5000CCA0321642D7] Id [145]
... All chunklets played back / relocated.
... cleared logging mode for cage 1 mag 13
... relocating chunklets from spare space

[Juliano_Flores]

Hello moinf,

This exactly situation (3PAR 7400: disk replaced + "servicemag resume -partial...") didn't work for us: the HPE was needed to replace this disk for a new one.

Best Regards,
Juliano Flores