DLPX-95197 NFS_OPS analytics collector not working

[sebroy]

See my latest comment in the Jira issue. While I was here, I made sure that the various copies of structure definitions for NFSv4 reads and writes accurately reflects the current kernel definitions.

Testing

ab-pre-push: https://selfservice-jenkins.eng-tools-prd.aws.delphixcloud.com/job/appliance-build-orchestrator-pre-push/12027/

estat:

root@ip-10-110-237-28:~# estat nfs 10
... (bcc warnings elided)
4 warnings generated.
08/27/25 - 17:23:15 UTC

 Tracing enabled... Hit Ctrl-C to end.
   microseconds                                          v4/write, async
value range                 count ------------- Distribution -------------
[300, 400)                      7 |@@@@
[400, 500)                     34 |@@@@@@@@@@@@@@@@@@@
[500, 600)                      2 |@
[600, 700)                      4 |@@
[700, 800)                      4 |@@
[800, 900)                      4 |@@
[900, 1000)                     2 |@
[1000, 2000)                    7 |@@@@
[2000, 3000)                    3 |@@
[3000, 4000)                    2 |@
[4000, 5000)                    1 |@

                                       iops(/s)  avg latency(us)       stddev(us)  throughput(k/s)
v4/write, async                               7              822           608038             7168


                                       iops(/s)  throughput(k/s)
total                                         7             7168

backend analytics:

hercules=> select * from analytics_datapoint where statistic_type='NFS_OPS';
 data_id | statistic_type |                                                                                           data                                                                                           |      timestamp      |    removal_time     | resolution |            slice
---------+----------------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+---------------------+---------------------+------------+-----------------------------
    2332 | NFS_OPS        | {"op":"write","avgLatency":3936765,"latency":{"4000000":1},"count":1,"throughput":1048576}                                                                                               | 2025-08-27 17:22:11 | 2025-08-27 23:22:11 |          1 | ANALYTICS_STATISTIC_SLICE-5
    2748 | NFS_OPS        | {"op":"write","avgLatency":1414027,"latency":{"500000":4,"800000":3,"3000000":2,"4000000":2,"1000000":2,"5000000":1,"2000000":4,"400000":1,"700000":1},"count":20,"throughput":20971520} | 2025-08-27 17:23:22 | 2025-08-27 23:23:22 |          1 | ANALYTICS_STATISTIC_SLICE-5
    2734 | NFS_OPS        | {"count":1,"op":"write","throughput":1048576,"avgLatency":3652440,"latency":{"4000000":1}}                                                                                               | 2025-08-27 17:22:00 | 2025-09-03 17:22:00 |         60 | ANALYTICS_STATISTIC_SLICE-5
    2720 | NFS_OPS        | {"count":1,"op":"write","throughput":1048576,"avgLatency":3936765,"latency":{"4000000":1}}                                                                                               | 2025-08-27 17:22:00 | 2025-09-03 17:22:00 |         60 | ANALYTICS_STATISTIC_SLICE-5
    2749 | NFS_OPS        | {"op":"write","avgLatency":581127,"latency":{"500000":12,"3000000":1,"900000":1,"2000000":1,"400000":4,"700000":1},"count":20,"throughput":20971520}                                     | 2025-08-27 17:23:23 | 2025-08-27 23:23:23 |          1 | ANALYTICS_STATISTIC_SLICE-5
    2733 | NFS_OPS        | {"op":"write","avgLatency":551273,"latency":{"500000":7,"2000000":1,"400000":1,"300000":1},"count":10,"throughput":10485760}                                                             | 2025-08-27 17:23:19 | 2025-08-27 23:23:19 |          1 | ANALYTICS_STATISTIC_SLICE-5
    2719 | NFS_OPS        | {"op":"write","avgLatency":3652440,"latency":{"4000000":1},"count":1,"throughput":1048576}                                                                                               | 2025-08-27 17:22:17 | 2025-08-27 23:22:17 |          1 | ANALYTICS_STATISTIC_SLICE-5
    2741 | NFS_OPS        | {"op":"write","avgLatency":618214,"latency":{"500000":5,"800000":1,"900000":1,"2000000":1,"600000":1,"700000":1},"count":10,"throughput":10485760}                                       | 2025-08-27 17:23:21 | 2025-08-27 23:23:21 |          1 | ANALYTICS_STATISTIC_SLICE-5
(8 rows)

[nealquigley]

As soon as an appliance-build finishes, I'll test this out with estat as well as making sure that the analytics backend is collecting NFS stats.

Looks good, modulo the above testing.