[pgpool-hackers: 2789] Re: Proposal: Disabling load balance after writing query

Tatsuo Ishii ishii at sraoss.co.jp
Mon Apr 23 13:32:54 JST 2018


Finally I have finished to implement this feature.
Full description of the feature is here:

http://tatsuo-ishii.github.io/pgpool-II/current/runtime-config-load-balancing.html#GUC-DISABLE-LOAD-BALANCE-ON-WRITE

Best regards,
--
Tatsuo Ishii
SRA OSS, Inc. Japan
English: http://www.sraoss.co.jp/index_en.php
Japanese:http://www.sraoss.co.jp

From: Tatsuo Ishii <ishii at sraoss.co.jp>
Subject: [pgpool-hackers: 2726] Proposal: Disabling load balance after writing query
Date: Fri, 16 Feb 2018 11:42:27 +0900 (JST)
Message-ID: <20180216.114227.1577486727559175532.t-ishii at sraoss.co.jp>

> Currently we have a feature to disable load balancing in a transaction
> if write query is sent. This allows legacy applications that do not
> consider streaming replication delay to be used with PostgreSQL's
> streaming replication. However, if the transaction closes, in the
> subsequent transaction load balancing starts again, which maybe not
> good for read query because of the replication delay.
> 
> So my proposal is, adding a new configuration parameter called
> "disable_load_balance_on_write" to allow more control on the existing
> feature. Here are possible values for the configuration:
> 
> off: completely disable prohibiting load balance feature, including
> existing behavior.
> 
> transaction: same as current behavior, i.e. if write query is sent,
> load balance is disabled within the transaction.
> 
> trans_transaction: the effect of "transaction" will continue in
> the subsequent transactions.
> 
> always: the effect will continue even out side of transactions.
> 
> Comments and suggestions are welcome.
> 
> Best regards,
> --
> Tatsuo Ishii
> SRA OSS, Inc. Japan
> English: http://www.sraoss.co.jp/index_en.php
> Japanese:http://www.sraoss.co.jp
> _______________________________________________
> pgpool-hackers mailing list
> pgpool-hackers at pgpool.net
> http://www.pgpool.net/mailman/listinfo/pgpool-hackers


More information about the pgpool-hackers mailing list