Re: Failure in subscription test 004_sync.pl

On 2021-Jun-14, Masahiko Sawada wrote:

> The condition should be the opposite; we should raise the error when
> 'nowait' is true. I think this is the cause of the test failure. Even
> if DROP SUBSCRIPTION tries to drop the slot with the WAIT option, we
> don't wait but raise the error.

Hi, thanks for diagnosing this and producing a patch! I ended up
turning the condition around, so that all three stanzas still test
"!nowait"; which seems a bit easier to follow.

TBH I'm quite troubled by the fact that this test only failed once on
each animal; they all had a lot of successful runs after that. I wonder
if this is because coverage is insufficient, or is it just bad luck.

I also wonder if this bug is what caused the random failures in the test
case I tried to add. I should look at that some more now ...

--
Álvaro Herrera Valdivia, Chile
Al principio era UNIX, y UNIX habló y dijo: "Hello world\n".
No dijo "Hello New Jersey\n", ni "Hello USA\n".