StorageGRID IO Errors Caused by Faulty SFP on SG5700 Appliance

Last updated
Save as PDF
Share
1. Share
2. Tweet
3. Share

Views:: 108

Visibility:: Public

Votes:: 0

Category:: storagegrid-webscale

Specialty:: sgrid

Last Updated:

Applies to

NetApp StorageGRID Appliance SG5700

Issue

IO errors causing the node to be unstable and experience Cassandra issues.

Check the E-series STATE-CAPTURE-DATA (by downloading a support bundle) and StorageGRID kern.log (by downloading a support bundle) on FC errors + IO errors, see if its matching the entries below to proceed with a replacement of the SFP's between de Compute and Storage controllers:

kern.log:

Feb 31 12:28:26 localhost kernel: [1127732.640909] sd 7:0:0:251: [sdaf] tag#191 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK
Feb 31 12:28:26 localhost kernel: [1127732.659173] print_req_error: I/O error, dev sdah, sector 6292015671
Feb 31 12:28:26 localhost kernel: [1127732.665596] device-mapper: multipath: Failing path 66:16.
Feb 31 12:28:27 localhost kernel: [1127733.370644] device-mapper: multipath: Reinstating path 66:48.

Under STATE-CAPTURE-DATA see if the following can be seen from the FCDump:

2806-A           Our  Num  ::...Exchange Counts...:: Num  ..Link Up..
Chip   LinkStat  Port Port ::                     :: Link  Bad  Bad
                  ID  Logi ::Open    Total  Errors:: Down  Char Frame
 2-Src Up-Ptp       2    2 ::  3   22429187      0::    8     0     0
 3-Src Up-Ptp       2    2 ::  6   15175346     500::    8     12277095     0