[pgpool-general: 5078] Re: Getting EOF encountered with frontend

[Yates, James C. -ND]

PgPool not responding has happened three times in the last 12 hours.  I turned off SSL a couple of hours ago, to see if that makes a difference.   I should be able to tell if turning off SSL makes a difference over night here in California.  When you say impolite behavior by clients, do you have anything specific?  This is a custom Ruby 2.0 and Ruby on Rails 4.1 application.

> On Oct 26, 2016, at 5:15 PM, Tatsuo Ishii <ishii at sraoss.co.jp> wrote:
> 
>> With PgPool 3.5.4 and Postgres 9.5.4.   I’m seeing ERROR(s) like the following, without SSL turned on:
>> 
>> 2016-10-26 16:01:27: pid 5896: ERROR:  unable to read data from frontend
>> 2016-10-26 16:01:27: pid 5896: DETAIL:  EOF encountered with frontend
>> 2016-10-26 16:01:27: pid 5896: LOCATION:  pool_stream.c:261
>> 
>> 
>> With SSL turned on I see the following:
>> 
>> 2016-10-25 10:30:26: pid 19075: WARNING:  ssl read: EOF detected
>> 2016-10-25 10:30:26: pid 19075: ERROR:  unable to read data from frontend
>> 2016-10-25 10:30:26: pid 19075: DETAIL:  socket read failed with an error "Success"
>> 
>> 
>> Eventually PgPool stops responding and I have to restart Pgpool.   This is a two node Master/Slave on Redhat EL 6.
>> 
>> I don’t see any errors in the Postgres Database log for these times,  PgPool is running on both of the Database Servers.
> 
> The error itself is caused by an impolite behavior of clients.
> However not responding Pgpool-II is not a good thing.
> 
> Do you have to restart Pgpool-II with/without SSL?
> 
> Best regards,
> --
> Tatsuo Ishii
> SRA OSS, Inc. Japan
> English: http://www.sraoss.co.jp/index_en.php
> Japanese:http://www.sraoss.co.jp