[
https://issues.apache.org/jira/browse/HBASE-25658?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17313269#comment-17313269
]
Duo Zhang commented on HBASE-25658:
---
Test scan again.
Master branch
{noformat}
Scan: 88711 row per
[
https://issues.apache.org/jira/browse/HBASE-25658?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17312932#comment-17312932
]
Duo Zhang commented on HBASE-25658:
---
The new result, no big difference between trace disabled and
[
https://issues.apache.org/jira/browse/HBASE-25658?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17312919#comment-17312919
]
Duo Zhang commented on HBASE-25658:
---
Oh, it seems that the problem is HDFS... After restarting all the
[
https://issues.apache.org/jira/browse/HBASE-25658?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17312822#comment-17312822
]
Duo Zhang commented on HBASE-25658:
---
You can see that, for scan and read, enabling tracing for every
[
https://issues.apache.org/jira/browse/HBASE-25658?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17312820#comment-17312820
]
Duo Zhang commented on HBASE-25658:
---
I set up a all-in-one jaeger receiver with downsampling 1.0,
[
https://issues.apache.org/jira/browse/HBASE-25658?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17300115#comment-17300115
]
Duo Zhang commented on HBASE-25658:
---
For write, the diff is ~2.4%
For scan, the diff is ~-5.9%
For
[
https://issues.apache.org/jira/browse/HBASE-25658?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17300111#comment-17300111
]
Duo Zhang commented on HBASE-25658:
---
Setup a cluster with 10 nodes, run PE tool for write, scan and
