Re: [PATCH] pgbench --throttle (submission 7 - with lag measurement)
От | Fabien COELHO |
---|---|
Тема | Re: [PATCH] pgbench --throttle (submission 7 - with lag measurement) |
Дата | |
Msg-id | alpine.DEB.2.02.1307142231060.3991@localhost6.localdomain6 обсуждение исходный текст |
Ответ на | Re: [PATCH] pgbench --throttle (submission 7 - with lag measurement) (Greg Smith <greg@2ndQuadrant.com>) |
Ответы |
Re: [PATCH] pgbench --throttle (submission 7 - with lag
measurement)
|
Список | pgsql-hackers |
Hello Greg, > Correct patch (and the little one from me again) attached this time. Please find an updated v15 with only comment changes: * The new comment about "is_throttled" was inverted wrt the meaning of the variable, at least to my understanding. * ISTM that the impact of the chosen 1000 should appear somewhere. * The trans_need_throttle comment was slightly wrong: the first transaction is also throttled, when initially entering doCustom. About 123456 12345 vs 123456.012345: My data parser is usually "gnuplot" or "my eyes", and in both cases the later option is better:-) About the little patch: Parameter "ok" should be renamed to something meaningful (say "do_finish"?). Also, it seems that when timer is exceeded in doCustom it is called with true, but maybe you intended that it should be called with false?? Moreover, under normal circumstances (throttling is significantly below the maximum rate) PQfinish will be called directly by threadRun while interrupting sleeping threads. Also, it is important to remove the connection because it serves as a marker to know whether a client must run or not. So basically I do not understand how it works. Note that it does not mean that it does not work, it just means that I do not really understand:-) -- Fabien.
В списке pgsql-hackers по дате отправления: