[pgpool-hackers: 3589] Re: [Proposal] add the process status of pgpool for extended query

Takuma Hoshiai hoshiai at sraoss.co.jp
Thu Apr 23 15:42:44 JST 2020


On Thu, 16 Apr 2020 14:52:34 +0900 (JST)
Tatsuo Ishii <ishii at sraoss.co.jp> wrote:

> Hi,
> 
> I have tested this patch.

Thank you for your test with patch.
 
> 2020-04-16 14:48:50: pid 10024: LOG:  DB node id: 0 backend pid: 10061 statement: Parse: BEGIN
> 2020-04-16 14:48:50: pid 10024: LOG:  DB node id: 0 backend pid: 10061 statement: B message
> 2020-04-16 14:48:50: pid 10024: LOG:  DB node id: 0 backend pid: 10061 statement: Execute: BEGIN
> 
> Is it possible to show the statement (in the case above "BEGIN") with B message?

Yes, Does it mean like following this?

2020-04-23 15:32:29: pid 11300: LOG:  DB node id: 0 backend pid: 11359 statement: Parse: BEGIN
2020-04-23 15:32:29: pid 11300: LOG:  DB node id: 1 backend pid: 11360 statement: Parse: BEGIN
2020-04-23 15:32:29: pid 11300: LOG:  DB node id: 0 backend pid: 11359 statement: Bind: BEGIN
2020-04-23 15:32:29: pid 11300: LOG:  DB node id: 1 backend pid: 11360 statement: Bind: BEGIN
2020-04-23 15:32:29: pid 11300: LOG:  DB node id: 0 backend pid: 11359 statement: Execute: BEGIN
2020-04-23 15:32:29: pid 11300: LOG:  DB node id: 1 backend pid: 11360 statement: Execute: BEGIN
2020-04-23 15:32:29: pid 11300: LOG:  DB node id: 0 backend pid: 11359 statement: C message
2020-04-23 15:32:29: pid 11300: LOG:  DB node id: 1 backend pid: 11360 statement: C message

I have attached a patch with fixed above.

By the way, this log is outputed by 'log_per_node_statement' parameter.


Best Regards,

Takuma Hoshiai


> 
> Best regards,
> --
> Tatsuo Ishii
> SRA OSS, Inc. Japan
> English: http://www.sraoss.co.jp/index_en.php
> Japanese:http://www.sraoss.co.jp
> 
> > Hi,
> > 
> > I have attached this patch to this email.
> > 
> > This patch is very simple. pgpool can display kind of query in process status
> > when the EXECUTE message is received from frontend.
> > 
> > In addition, this patch will fix that log_statement support extended query.
> > Currently, log_statement only support a simple query.
> > 
> > Comments and suggestions are welcome.
> > 
> > 
> > Best Regards,
> > 
> > Takuma Hoshiai
> > 
> > On Thu, 17 Oct 2019 13:52:18 +0900
> > Takuma Hoshiai <hoshiai at sraoss.co.jp> wrote:
> > 
> >> Hi,
> >> 
> >> The pgpool’s prosess status display kind of query what is executing. 
> >> 
> >> e.g.
> >>  pgpool: hoshiai test  SELECT
> >>  pgpool: hoshiai test  END;
> >>  pgpool: hoshiai test  UPDATE
> >>  pgpool: hoshiai test  idle in transaction
> >>  pgpool: hoshiai test  idle in transaction
> >> 
> >> But it is only supported simple query now. When extended query is executed, 
> >> prosess status is 'idle' or 'idle in transaction' even if query is executing.
> >>  
> >> This status is incorrect and user misread status. 
> >> So in pgpool 4.2, I think new other status must display during extended query.
> >> 
> >> 
> >> Best Regards,
> >> 
> >> -- 
> >> Takuma Hoshiai <hoshiai at sraoss.co.jp>
> >> 
> >> _______________________________________________
> >> pgpool-hackers mailing list
> >> pgpool-hackers at pgpool.net
> >> http://www.pgpool.net/mailman/listinfo/pgpool-hackers
> >> 
> > 
> > 
> > -- 
> > Takuma Hoshiai <hoshiai at sraoss.co.jp>


-- 
Takuma Hoshiai <hoshiai at sraoss.co.jp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: fix_process_status_v2.patch
Type: application/octet-stream
Size: 2071 bytes
Desc: not available
URL: <http://www.sraoss.jp/pipermail/pgpool-hackers/attachments/20200423/18630224/attachment.obj>


More information about the pgpool-hackers mailing list