Postgres Pro
Downloads
Home   >   mailing lists

Thread: Is Backgroundworker.bgw_restart_time is defined in seconds?

Is Backgroundworker.bgw_restart_time is defined in seconds?

From
constzl
Date:

Hi,

typedef struct BackgroundWorker
{
    char        bgw_name[BGW_MAXLEN];
    char        bgw_type[BGW_MAXLEN];
    int         bgw_flags;
    BgWorkerStartTime bgw_start_time;
    int         bgw_restart_time;   /* in seconds, or BGW_NEVER_RESTART */
    char        bgw_library_name[BGW_MAXLEN];
    char        bgw_function_name[BGW_MAXLEN];
    Datum       bgw_main_arg;
    char        bgw_extra[BGW_EXTRALEN];
    pid_t       bgw_notify_pid; /* SIGUSR1 this backend on start/stop */
} BackgroundWorker;


static bool
SanityCheckBackgroundWorker(BackgroundWorker *worker, int elevel)
{
... ...

    if ((worker->bgw_restart_time < 0 &&
         worker->bgw_restart_time != BGW_NEVER_RESTART) ||
        (worker->bgw_restart_time > USECS_PER_DAY / 1000))

> The Backgroundworker.bgw_restart_time is defined in seconds,
> so should it be "USECS_PER_DAY / 1000,000" here instead of "USECS_PER_DAY / 1000"?
> Or am I getting it wrong?

    {
        ereport(elevel,
                (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
                 errmsg("background worker \"%s\": invalid restart interval",
                        worker->bgw_name)));
        return false;
    }
... ...
}

Regards!

Re: Is Backgroundworker.bgw_restart_time is defined in seconds?

From
Daniel Gustafsson
Date:
> On 26 Aug 2021, at 09:38, constzl <const_sunny@126.com> wrote:

>     if ((worker->bgw_restart_time < 0 &&
>          worker->bgw_restart_time != BGW_NEVER_RESTART) ||
>         (worker->bgw_restart_time > USECS_PER_DAY / 1000))
>
> > The Backgroundworker.bgw_restart_time is defined in seconds,
> > so should it be "USECS_PER_DAY / 1000,000" here instead of "USECS_PER_DAY / 1000"?
> > Or am I getting it wrong?

bgw_restart_time is defined as "It can be any positive value” in the docs, see:

    https://www.postgresql.org/docs/current/bgworker.html

My reading of the above code is that is tries to catch nonsensical values, not
to cap it to enforce restarts within a 24h period.

--
Daniel Gustafsson        https://vmware.com/

Re:Re: Is Backgroundworker.bgw_restart_time is defined in seconds?

From
constzl
Date:
Yes, I got it wrong. You're right. Thank you for your explanation!<br/><br/>Best Regards,
At 2021-08-26 16:09:36, "Daniel Gustafsson" <daniel@yesql.se> wrote:
>> On 26 Aug 2021, at 09:38, constzl <const_sunny@126.com> wrote:
>
>>     if ((worker->bgw_restart_time < 0 &&
>>          worker->bgw_restart_time != BGW_NEVER_RESTART) ||
>>         (worker->bgw_restart_time > USECS_PER_DAY / 1000))
>> 
>> > The Backgroundworker.bgw_restart_time is defined in seconds,
>> > so should it be "USECS_PER_DAY / 1000,000" here instead of "USECS_PER_DAY / 1000"?
>> > Or am I getting it wrong?
>
>bgw_restart_time is defined as "It can be any positive value” in the docs, see:
>
>    https://www.postgresql.org/docs/current/bgworker.html
>
>My reading of the above code is that is tries to catch nonsensical values, not
>to cap it to enforce restarts within a 24h period.
>
>--
>Daniel Gustafsson        https://vmware.com/
>
>

Re: Is Backgroundworker.bgw_restart_time is defined in seconds?

From
Tom Lane
Date:
Daniel Gustafsson <daniel@yesql.se> writes:
>> On 26 Aug 2021, at 09:38, constzl <const_sunny@126.com> wrote:
>> if ((worker->bgw_restart_time < 0 &&
>>      worker->bgw_restart_time != BGW_NEVER_RESTART) ||
>>     (worker->bgw_restart_time > USECS_PER_DAY / 1000))

> My reading of the above code is that is tries to catch nonsensical values, not
> to cap it to enforce restarts within a 24h period.

This may be intended to guard against integer overflows later (he guesses
without having read the code).

            regards, tom lane