[
https://issues.apache.org/jira/browse/KUDU-2986?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17015578#comment-17015578
]
Alexey Serbin commented on KUDU-2986:
-------------------------------------
[~helifu], did you try to repro
[this|https://issues.apache.org/jira/browse/KUDU-2986?focusedCommentId=16961668&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-16961668]
scenario with the commit caf2b50529ce59bb4847e76c3e32f2574bae7809 ?
I'm still getting inconsistent results reported by {{kudu}} CLI on both space
used and number of rows.
With kudu-master and one kudu-tserver of version 1.12.x (the snapshot was at
{{c235ba188d282ebba6c83e19a0aab07846d4b91a}} chagelist, so it included
{{caf2b50529ce59bb4847e76c3e32f2574bae7809}} fix), two kudu-tservers of version
1.10.1, I used the following command to create a table:
{noformat}
./bin/kudu perf loadgen --table_num_replicas=3 --table_num_range_partitions=2
--keep-auto-table 10.17.240.17:9876
{noformat}
{noformat}
$ ./bin/kudu table statistics 10.17.240.17:9876
default.loadgen_auto_1441c9c27158421ab3f2c1ee62e3db8c
TABLE default.loadgen_auto_1441c9c27158421ab3f2c1ee62e3db8c
on disk size: 25236567
live row count: 346
{noformat}
With kudu-master and kudu-tserver started from binaries of 1.12.x snapshot at
{{c235ba188d282ebba6c83e19a0aab07846d4b91a}} changelist:
{noformat}
$ ./bin/kudu table statistics 10.17.240.17:9876
default.loadgen_auto_1441c9c27158421ab3f2c1ee62e3db8c
TABLE default.loadgen_auto_1441c9c27158421ab3f2c1ee62e3db8c
on disk size: 134760550
live row count: 2000
{noformat}
Do I miss something?
> Incorrect value for the 'live_row_count' metric with pre-1.11.0 tables
> ----------------------------------------------------------------------
>
> Key: KUDU-2986
> URL: https://issues.apache.org/jira/browse/KUDU-2986
> Project: Kudu
> Issue Type: Bug
> Components: CLI, client, master, metrics
> Affects Versions: 1.11.0
> Reporter: YifanZhang
> Priority: Major
>
> When we upgraded the cluster with pre-1.11.0 tables, we got inconsistent
> values for the 'live_row_count' metric of these tables:
> When visiting masterURL:port/metrics, we got 0 for old tables, and got a
> positive integer for a old table with a newly added partition, which is the
> count of rows in the newly added partition.
> When gettingĀ table statistics via `kudu table statistics` CLI tool, we got 0
> for old tables and the old table with a new parition.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)