Hi question about disk labels we | NetApp | Page 1

sinful mantle Dec 15, 2022, 8:17 PM

#

hi @next harness how long have you waited? Wonder if Prometheus is reporting the unassigned and the assigned until the unassigned expires? Can you check in Prometheus and see if both are returned? Looking at the disk template, uniqueness is determined by disk-uid and disk-name. I believe the replaced disk will have a different uid which means Prometheus may have both instances for a bit

next harness Dec 15, 2022, 8:20 PM

#

it's one hour now after assigning ownership

sinful mantle Dec 15, 2022, 8:25 PM

#

thanks and if you check Prom do you see the failed disk and the new disk?

next harness Dec 15, 2022, 8:25 PM

#

disk_labels returns 24 metrics and that's the total number of disks in the cluster.

#

disk_labels{cluster="flc1-sys-prt1-coresys", container_type="shared", datacenter="PRT", disk="2.0.19", failed="false", firmware_revision="NA02", instance="flc1-sys-prt1-coresys", job="netapp-harvest", model="X341_SSKBE900A10", node="flc1-01-sys-prt1-coresys", outage="unassigned", owner_node="flc1-01-sys-prt1-coresys", serial_number="WFKD03L4", shared="true", shelf="7377188709118577232", shelf_bay="19", type="SAS"}

#

the serial number is the same from ontap command disk show -disk 2.0.19

#

we will wait and see if it will clear

#

Disk     Partition Home              Owner             Home ID     Owner ID
-------- --------- ----------------- ----------------- ----------- -----------
2.0.19   Container flc1-01-sys-prt1-coresys flc1-01-sys-prt1-coresys 538015558 538015558
         Root      flc1-01-sys-prt1-coresys flc1-01-sys-prt1-coresys 538015558 538015558
         Data      flc1-01-sys-prt1-coresys flc1-01-sys-prt1-coresys 538015558 538015558```

sinful mantle Dec 15, 2022, 8:30 PM

#

what if you throw in a -fields outage-reason on the CLI?

next harness Dec 15, 2022, 8:31 PM

#

empty

disk   outage-reason                                                  
------ ------------- 
2.0.19 -```

#

somehow it is stucked in prom or harvest?

sinful mantle Dec 15, 2022, 8:33 PM

#

right I was going to say, the other possibility is ONTAP is still returning that record with an outage reason and Harvest is dutifully reporting it? we can skip Prometheus and send the same ZAPI or REST request to ONTAP directly to rule that out if you want.

next harness Dec 15, 2022, 8:34 PM

#

let me restart harvest first

#

gone after restarting harvest

sinful mantle Dec 15, 2022, 8:36 PM

#

well phooey

#

i'll open an issue so we can try to recreate. Sounds like a Harvest bug. Thanks for raising!

next harness Dec 15, 2022, 8:38 PM

#

No problem! Thank you for the great support !

sinful mantle Dec 15, 2022, 8:40 PM

#

created https://github.com/NetApp/harvest/issues/1626

#Hi question about disk labels we