Monitoring IO in Vertica
Hello Everyone,
I'm new to Vertica and currently setting up some alerts to monitor utilization of our cluster. One thing I noticed at this point is there are no I/O data being generated on the dc_io_info_by_minute table.
dbadmin=> SELECT dbadmin-> node_name, dbadmin-> round(min(start_time), 'MI') as start_time, dbadmin-> round(max(end_time), 'MI') as end_time, dbadmin-> round( ( (sum(total_read_bytes_end_value - total_read_bytes_start_value) / 1024) / dbadmin(> (datediff('millisecond', min(start_time), max(end_time)) / 1000.0) )::float, 2) as read_kbytes_per_sec, dbadmin-> round( ( (sum(total_written_bytes_end_value - total_written_bytes_start_value) / 1024) / dbadmin(> (datediff('millisecond', min(start_time), max(end_time)) / 1000.0) )::float, 2) as written_kbytes_per_sec dbadmin-> FROM v_internal.dc_io_info_by_minute dbadmin-> GROUP BY node_name, round(start_time, 'MI') dbadmin-> ORDER BY 3 desc dbadmin-> limit 10; node_name | start_time | end_time | read_kbytes_per_sec | written_kbytes_per_sec ---------------------+---------------------+---------------------+---------------------+------------------------ v_medicaid_node0001 | 2017-02-24 03:24:00 | 2017-02-24 03:25:00 | 0 | 0 v_medicaid_node0001 | 2017-02-24 03:23:00 | 2017-02-24 03:24:00 | 0 | 0 v_medicaid_node0001 | 2017-02-24 03:22:00 | 2017-02-24 03:23:00 | 0 | 0 v_medicaid_node0001 | 2017-02-24 03:21:00 | 2017-02-24 03:22:00 | 0 | 0 v_medicaid_node0001 | 2017-02-24 03:20:00 | 2017-02-24 03:21:00 | 0 | 0 v_medicaid_node0001 | 2017-02-24 03:19:00 | 2017-02-24 03:20:00 | 0 | 0 v_medicaid_node0001 | 2017-02-24 03:18:00 | 2017-02-24 03:19:00 | 0 | 0
At the same time I used iostat to check the utilization at the OS level.
Device: rrqm/s wrqm/s r/s w/s rkB/s wkB/s avgrq-sz avgqu-sz await r_await w_await svctm %util sdd 0.00 0.28 0.13 8.84 15.99 265.52 62.81 0.19 21.73 12.30 21.87 0.52 0.46 sdc 0.00 8.86 0.00 0.68 0.01 182.90 532.88 0.18 256.43 5.44 257.21 4.16 0.29 sde 0.00 0.01 0.43 0.29 145.86 146.18 811.76 0.10 142.53 8.50 336.57 3.92 0.28 sda 0.00 0.03 0.01 0.45 0.28 2.18 10.77 0.00 2.78 5.99 2.73 1.51 0.07 sdb 0.00 0.01 0.41 0.34 140.90 146.32 763.32 0.23 301.64 6.93 657.27 5.77 0.43 dm-0 0.00 0.00 0.00 0.00 0.00 0.00 8.00 0.00 0.55 0.55 0.00 0.42 0.00 dm-1 0.00 0.00 0.00 0.00 0.05 0.02 28.31 0.00 5.19 5.72 4.87 1.95 0.00 dm-2 0.00 0.00 0.01 0.28 0.21 1.32 10.70 0.00 2.66 6.28 2.60 1.30 0.04 dm-3 0.00 0.00 0.00 0.00 0.00 0.02 26.01 0.00 12.20 1.70 13.17 2.05 0.00 dm-4 0.00 0.00 0.00 0.11 0.00 0.51 9.02 0.00 2.63 1.91 2.63 1.56 0.02 dm-5 0.00 0.00 0.00 0.03 0.00 0.07 4.79 0.00 2.93 2.90 2.93 2.14 0.01 dm-6 0.00 0.00 0.00 0.00 0.00 0.00 30.92 0.00 1.95 1.94 2.06 1.84 0.00 dm-7 0.00 0.00 0.97 19.32 302.76 740.92 102.87 0.76 37.68 8.33 39.15 0.65 1.32 dm-8 0.00 0.00 0.00 0.05 0.02 0.23 9.26 0.00 3.19 4.97 3.18 1.41 0.01
And here is the collector policy
`dbadmin=> SELECT get_data_collector_policy('IoInfoByMinute');
get_data_collector_policy
64KB kept in memory, 25600KB kept on disk. Time based retention disabled.
(1 row)
`
Is there something I'm missing here? We are using Vertica 7.2. Thanks in advance!
Comments
would appreciate any advice/input on the issue above. Thanks!