Re: Long text values destroys logical replication slots

<div dir="ltr">I managed to reproduce this problem on postgres 9.4.1 and 9.4.5 on osx and gentoo.</div><div
class="gmail_extra"><br/><div class="gmail_quote">On Wed, Oct 28, 2015 at 9:54 PM, Andres Freund <span dir="ltr"><<a
href="mailto:andres@anarazel.de"target="_blank">andres@anarazel.de</a>></span> wrote:<br /><blockquote
class="gmail_quote"style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="HOEnZb"><div
class="h5">OnOctober 28, 2015 7:51:18 PM GMT+01:00, "Adam Dratwiński" <<a
href="mailto:adam.dratwinski@gmail.com">adam.dratwinski@gmail.com</a>>wrote:<br /> >I recently encountered a bug
onmy production postgres which results in<br /> >killing replication slot. It happens when there is an UPDATE
statement<br/> >called on a record having text column with a very long value (about<br /> >1mb).<br /> >The
problemcan be easily reproduced by this query:<br /> ><br /> ><a
href="https://gist.github.com/adwinsky/f7a6dc4382776fb9ea60"rel="noreferrer"
target="_blank">https://gist.github.com/adwinsky/f7a6dc4382776fb9ea60</a><br/> ><br /> >Here I am using
decoderbufsdecoder<br /> ><a href="https://github.com/krzychk/decoderbufs" rel="noreferrer"
target="_blank">https://github.com/krzychk/decoderbufs</a><br/> >but it doesn't seems to be related with this plugin
asit also happen<br /> >on<br /> >the postgres test decoder.<br /><br /></div></div>Did you try the last minor
release?<br/><br /> Andres<br /> Hi<br /> ---<br /> Please excuse brevity and formatting - I am writing this on my
mobilephone.<br /></blockquote></div><br /></div>