On Tue, May 7, 2019 at 9:51 AM Tom Lane <t...@sss.pgh.pa.us> wrote: > > Amit Kapila <amit.kapil...@gmail.com> writes: > > On Tue, May 7, 2019 at 9:44 AM Tom Lane <t...@sss.pgh.pa.us> wrote: > >> Shouldn't there be a catversion bump in this? > > > Won't the next vacuum run will create the FSM? > > If that's true, and if we don't throw any errors meanwhile, it's fine. >
This is what is suppose to happen. > I'm just worried that that's an untested scenario. > I can test it. Basically, without revert of the patch, I can create a table with a few records (so as to ensure that there is no FSM). And then revert the patch and upgrade the database and perform few dml operations on that table and see what happens? -- With Regards, Amit Kapila. EnterpriseDB: http://www.enterprisedb.com
