<div dir="auto">Hi, <div dir="auto"><br></div><div dir="auto">We've been running the patch on our testing environment for almost a week now and haven't seen a single authentication error since the fix. Before the patch we saw multiple errors on most days. So, I'd say the issue is fixed. </div><div dir="auto"><br></div><div dir="auto">Best regards, </div><div dir="auto">Emond </div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">Op vr 25 okt 2024 03:17 schreef Tatsuo Ishii <<a href="mailto:ishii@postgresql.org">ishii@postgresql.org</a>>:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">> Op do 24 okt 2024 om 07:53 schreef Tatsuo Ishii <<a href="mailto:ishii@postgresql.org" target="_blank" rel="noreferrer">ishii@postgresql.org</a>>:<br>
> <br>
>> Hi Emond,<br>
>> > In pool_get_user_credentials in pool_passwd.c (which it must reach<br>
>> because<br>
>> > health_check_password is empty), the password file is read via passwd_fd.<br>
>> > This file descriptor is rewound at line 362 (in 4.5.4). However, I wonder<br>
>> > what would happen if another process uses this same descriptor at the<br>
>> same<br>
>> > time. I suspect this might lead to one of the two processes (or maybe<br>
>> even<br>
>> > both) to get incorrect results.<br>
>><br>
>> I think your analythis is correct. Yes, if multiple process are<br>
>> reading from the same file descriptor, it could lead to confusing<br>
>> results. Currently pgpool main process opens the pool_passwd file,<br>
>> then health check process inherit the file descriptor. This is not<br>
>> good. Attached is a patch to let the health check process to reopen<br>
>> the pool_passwd file so that it has it's own file descriptor.<br>
>><br>
> <br>
> Thank you for the quick response and the patch. I've deployed a patched<br>
> version of pgpool onto a testing environment, which we will monitor for the<br>
> upcoming week. Until we've verified the patch actually fixes the issue,<br>
> we'll use the health_check_password setting as a workaround.<br>
<br>
OK, I'm looking forward to hearing from you about the testing result.<br>
<br>
Best reagards,<br>
--<br>
Tatsuo Ishii<br>
SRA OSS K.K.<br>
English: <a href="http://www.sraoss.co.jp/index_en/" rel="noreferrer noreferrer" target="_blank">http://www.sraoss.co.jp/index_en/</a><br>
Japanese:<a href="http://www.sraoss.co.jp" rel="noreferrer noreferrer" target="_blank">http://www.sraoss.co.jp</a><br>
</blockquote></div>