Comment 8 for bug 1999711

I noticed a follow-up commit in the upstream git repo a week or so after the original fix:

original fix:
commit 62b2907a46c9e495c1c71114cc8f18ed137b7f34
Author: Chengen Du <email address hidden>
Date: Wed Nov 23 14:11:36 2022 +0800

    apps/snmptrap: Fix mysql reconnection after hitting wait_timeout

    MySQL v8.0 supports ER_CLIENT_INTERACTION_TIMEOUT, which is not handled
    in snmptrapd

    Signed-off-by: Chengen Du <email address hidden>

A few days later, this changes "bool" to "int":
commit a1c80f7778da0aaadfb3fc44b213586f632d8d5d
Author: Bart Van Assche <email address hidden>
Date: Mon Dec 5 19:19:40 2022 -0800

    apps/snmptrapd_sql: Fix a recently introduced build error

    Fixes: 62b2907a46c9 ("apps/snmptrap: Fix mysql reconnection after hitting wait_timeout")

I vaguely remember a mysql build issue when mysql 8 came along, where mysql stopped exporting my_bool iirc. I see we use "bool" all the way to focal, and then in bionic use "int". And the patch correctly flags that bionic change as being a backport.

Can you detail the test case a bit more, for those not familiar setting up snmp traps? I'll sponsor lunar, but a more detailed step by step test case for the SRU would be great.