[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-08-19 Thread Uwe L. Korn (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16585094#comment-16585094 ] Uwe L. Korn commented on ARROW-2059: Also adding {{-fno-omit-frame-pointer}} makes the stacks much

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-08-06 Thread Wes McKinney (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16570906#comment-16570906 ] Wes McKinney commented on ARROW-2059: - Ah great, I will try that out > [Python] Possible performance

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-08-06 Thread Antoine Pitrou (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16569934#comment-16569934 ] Antoine Pitrou commented on ARROW-2059: --- If you get a lot of unknown symbols in the call stacks, my

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-08-03 Thread Wes McKinney (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16568657#comment-16568657 ] Wes McKinney commented on ARROW-2059: - I just attached a FlameGraph (perf.svg) created after the work

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-07-16 Thread Wes McKinney (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16545499#comment-16545499 ] Wes McKinney commented on ARROW-2059: - Oh great, thanks =) Well I'll run some perf tests locally,

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-07-16 Thread Antoine Pitrou (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16545492#comment-16545492 ] Antoine Pitrou commented on ARROW-2059: --- The patch was already merged actually (see

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-07-16 Thread Wes McKinney (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16545484#comment-16545484 ] Wes McKinney commented on ARROW-2059: - Thanks [~pitrou] for investigating. I agree there doesn't seem

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-07-16 Thread Antoine Pitrou (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16545417#comment-16545417 ] Antoine Pitrou commented on ARROW-2059: --- There doesn't seem to be anything obvious left to do here.

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-07-12 Thread Antoine Pitrou (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16541928#comment-16541928 ] Antoine Pitrou commented on ARROW-2059: --- I use Linux "perf" on a benchmark script than ran "perf

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-07-12 Thread Wes McKinney (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16541910#comment-16541910 ] Wes McKinney commented on ARROW-2059: - Thanks [~pitrou], do you have some tricks for creating

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-07-12 Thread Antoine Pitrou (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16541898#comment-16541898 ] Antoine Pitrou commented on ARROW-2059: --- Here is a profile tree got by exercising the write path on

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-02-05 Thread Wes McKinney (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16352638#comment-16352638 ] Wes McKinney commented on ARROW-2059: - Thanks [~alphalfalfa] -- this suggests that we need to optimize

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-02-04 Thread Jingyuan Wang (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16352025#comment-16352025 ] Jingyuan Wang commented on ARROW-2059: -- Stack trace of the failure when processing 100M-row data

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-02-04 Thread Jingyuan Wang (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16352024#comment-16352024 ] Jingyuan Wang commented on ARROW-2059: -- Stack trace of the failure when processing 100M-row data

[jira] [Commented] (ARROW-2059) [Python] Possible performance regression in Feather read/write path

2018-02-04 Thread Jingyuan Wang (JIRA)
[ https://issues.apache.org/jira/browse/ARROW-2059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16352023#comment-16352023 ] Jingyuan Wang commented on ARROW-2059: -- Here is what I've done. I simple repeated the 1M rows and