Monitoring NFS/SOFS Ingestion

Zenko implements an NFS feature that ingests the NFS file hierarchy that the RING’s Scale-Out File System (SOFS) projects. For this release, the only available metrics for NFS/SOFS ingestion operations reside in log files. SOFS metadata ingestion runs on a cron schedule, and users can make a query against Cosmos on the kubectl location to discover how and when scheduled NFS activities have been executed.

To find the relevant Cosmos pods, enter:

$ kubectl get pods  | grep "rclone"

nfs-options-cosmos-rclone-1555329600-7fr27                    0/1     Completed   0          7h
nfs-options-cosmos-rclone-1555333140-bb5vf                    1/1     Running     0          1h

Use this information to retrieve the logs with a command formatted as follows:

$ kubectl logs -f {{location}}-cosmos-rclone-{{hash}}

Hence:

$ kubectl logs -f my-location-name-cosmos-rclone-84988dc9d4-9dwjl

yields:

* myobjects273725
* myobjects273726
* myobjects273727

2019/04/15 17:48:30 INFO  : S3 bucket nfs-rsize: Waiting for checks to finish
2019/04/15 17:48:31 INFO  : S3 bucket nfs-rsize: Waiting for transfers to finish
2019/04/15 17:48:31 INFO  : Waiting for deletions to finish
2019/04/15 17:48:31 INFO  :
Transferred:                0 / 0 Bytes, -, 0 Bytes/s, ETA -
Errors:                 0
Checks:            500000 / 500000, 100%
Transferred:            0 / 0, -
Elapsed time:  2h48m18.8s

Response	Description
Transferred	The byte count of metadata transferred, with data rate in B/s and a completion time estimate. Transferred counts are almost always 0/0.
Errors	Aggregated error count for the requested job.
Checks	Ingested information/total information to be ingested, and a percentage expressing this ratio.
Elapsed Time	Time spent on the current ingestion cycle.