Identify device

Hi
We're looking for a device. All I have is these lines, and I'd like to find it so we can replace it. Audit log says it's failing.
It's a Tegile HD2080 SAN.
Thanks in advance
Gabe

audit.log:
Code:
Description: NVDIMM device '/pci@0,0/pci8086,2f08@3/pci10b5,8717@0/pci10b5,8717@1/pci10b5,87b0@0/ntb_transport@0/pram@0' is unusable: State transition ARMED -> UNARMED.
Details: <html><body><b>NVL6500E65000</b><br/>NVDIMM device '/pci@0,0/pci8086,2f08@3/pci10b5,8717@0/pci10b5,8717@1/pci10b5,87b0@0/ntb_transport@0/pram@0' is unusable: State transition ARMED -> UNARMED.<br/><br/><b>Event Information:  </b><br/>1) Host Name: TegileHD2080-B<br/>2) NVDIMM Device Path: /pci@0,0/pci8086,2f08@3/pci10b5,8717@0/pci10b5,8717@1/pci10b5,87b0@0/ntb_transport@0/pram@0<br/><br/><b>User</b>:  SYSTEM</body></html>
lshal > hwlist.txt:
Code:
udi = '/org/freedesktop/Hal/devices/pci_0_0/pci8086_2f08_3/pci10b5_8717_0/pci10b5_8717_1/pci10b5_87b0_0/ntb_transport_0/pram_0/ppram_copy_0/nvram_0_0'
  info.solaris.driver = 'nvram'  (string)
  solaris.devfs_path = '/pci@0,0/pci8086,2f08@3/pci10b5,8717@0/pci10b5,8717@1/pci10b5,87b0@0/ntb_transport@0/pram@0/ppram_copy@0/nvram@0'  (string)
  info.product = 'nvram'  (string)
  info.udi = '/org/freedesktop/Hal/devices/pci_0_0/pci8086_2f08_3/pci10b5_8717_0/pci10b5_8717_1/pci10b5_87b0_0/ntb_transport_0/pram_0/ppram_copy_0/nvram_0_0'  (string)
  info.parent = '/org/freedesktop/Hal/devices/pci_0_0/pci8086_2f08_3/pci10b5_8717_0/pci10b5_8717_1/pci10b5_87b0_0/ntb_transport_0/pram_0/ppram_copy_0_0'  (string
 
Hi,

a) that's Solaris

b) Tegile used SMART Memory NVDIMM-N modules. Yours seems to have had a 8GB NVDIMM.
You find more info in my notes at https://github.com/FlorianHeigl/Tegile_Storage_Info/


There's usually two types, JEDEC and pre-JEDEC, my Tegile had the pre-jedec model though I have the suspicion that both work if you have a sufficiently recent bios. Maybe it's also the battery that is failing.
Your part SHOULD be the one mentioned here:
Try to do it proper and get the SuperMicro branded one so you can ask for some help from them maybe.

You could look at
/usr/bin/nvadm status
and
/usr/bin/nvadm health
to find out more. Check in both nodes.
I think there was a command that outputs
smartctl -a
style text for the nvdimms to make it easier to understand.

From a technical stance, there's 3 components that could be meant to be failing: DIMM - memory errors, that's fatal. Flash - that would mean it can no longer savely commit the data on shutdown, that's fatal. Battery ("PowerGem" for other vendors) - that's something that can be replaced, with some care. But the battery still is a whole module, not just a normal battery. Best is to get the whole assemble with the DIMM.
My advice get one of the unused spares kits that are on ebay, it'll contain equal or even perfectly matching spares for most parts you could need.
 
Back
Top