VMware Cloud Community
vanree
Enthusiast
Enthusiast

Coredump partition disappeared after reboot

Hello,

We have 2 Dell R320 boxes with ESX 5.5 without disks, the systems boot from the dual SD cards fine and last week I found this article on Dell website:

Configuring iDRAC vFlash as a VMware ESXi VMKernel Coredump Collector - White Papers - TechCenter E...

Since we have the iDrac7 with its own 8GB SD card, I thought that is a good idea to configure the LOG and COREDUMP onto.

What this basically show us in that article is to turn the USBarbitrator off permanently and to create a partition on the iDrac SD card, which shows up as a disk in ESXi.

It worked fine until I rebooted a host.

After reboot  the LOG partition still works fine, but the COREDUMP partition is gone according to ESXi 5.5.

Also vCenter complains for this host that the coredump target has not been configured.

SSH into the box and checked the coredump partition is still there:

~ # fdisk /dev/disks/mpx.vmhba33\:C0\:T0\:L1 -l

***

*** The fdisk command is deprecated: fdisk does not handle GPT partitions.  Please use partedUtil

***

Disk /dev/disks/mpx.vmhba33:C0:T0:L1: 838 MB, 838860800 bytes

128 heads, 1 sectors/track, 12800 cylinders

Units = cylinders of 128 * 512 = 65536 bytes

                           Device Boot      Start         End      Blocks  Id System

/dev/disks/mpx.vmhba33:C0:T0:L1p1               1       12800      819199+ fc VMKcore

Even the partition type is correct (fc).

But then I do the coredump partition list, which normally showed this partition, now it's gone:

~ # esxcli system coredump partition list

no configured dump partition found; skipping

Been searching and looking, any ideas?

Thanks, Edwin

2 Replies
vanree
Enthusiast
Enthusiast

An update on this issue. I could not find why it did not see the coredump partition, so in the end I redid the cli commands to enable the location for coredump again and ESXi was happy again. Not sure if it will stick after reboot, I will report again if it does not.

The commands used were (whereby the partition location is the iDrac SD card partition, which might be different in your server):

esxcli system coredump partition set --partition="mpx.vmhba33:C0:T0:L1:1"

esxcli system coredump partition set --enable=true

niam
Enthusiast
Enthusiast

We have same issues here. But not not after a normal reboot, but after Update Manager installs new patches and reboots!

With command partedUtil you can see your coredump partition:

partedUtil getptbl /dev/disks/mpx.vmhba33\:C0\:T0\:L0

gpt

969 255 63 15572992

1 64 8191 C12A7328F81F11D2BA4B00A0C93EC93B systemPartition 128

5 8224 520191 EBD0A0A2B9E5443387C068B6B72699C7 linuxNative 0

6 520224 1032191 EBD0A0A2B9E5443387C068B6B72699C7 linuxNative 0

7 1032224 1257471 9D27538040AD11DBBF97000C2911D1B8 vmkDiagnostic 0

8 1257504 1843199 EBD0A0A2B9E5443387C068B6B72699C7 linuxNative 0

With vanree's answer you can repair it:

esxcli system coredump partition set --partition="mpx.vmhba33:C0:T0:L0:7"

esxcli system coredump partition set --enable=true

esxcli system coredump partition list

Name Path Active  Configured

---------------------- ------------------------------------------  ------  ----------

  1. mpx.vmhba33:C0:T0:L0:7 /vmfs/devices/disks/mpx.vmhba33:C0:T0:L0:7 true        true
0 Kudos