Re: Re: BUG #12990: Missing pg_multixact/members files (appears to have wrapped, then truncated)

On Wed, May 6, 2015 at 3:56 PM, Thomas Munro <thomas(dot)munro(at)enterprisedb(dot)com>
wrote:
>
> On Wed, May 6, 2015 at 2:29 PM, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:

Few comments:

1.
+ /*
+ * Override the multixact freeze settings if we are running out of
+ * member address space.
+ */
+ if (safe_multixact_age >= 0)
+ {
+ multixact_freeze_table_age = Min(safe_multixact_age,
+ multixact_freeze_table_age);

+ /* Special settings if we are running out of member address space. */
+ if (safe_multixact_age >= 0)
+ multixact_freeze_max_age = Min(multixact_freeze_max_age,
safe_multixact_age);
+

Some places use safe_multixact_age as first parameter and some
places use it at second place. I think it is better to use in same
order for the sake of consistency.

2.
in the hope
+ * that different tables will be vacuumed at different times due to their
+ * varying relminmxid values.

Does above line in comment on top of MultiXactCheckMemberUsage()
makes much sense?

3.
+ * we know the age of the oldest multixact in the system, so that's the
+ * value we want to when members is near safe_member_count. It should

typo.
so that's the value we want to *use* when ..

With Regards,
Amit Kapila.
EnterpriseDB: http://www.enterprisedb.com