On Fri, Jul 29, 2016 at 3:18 PM, Condor <condor@xxxxxxxxxx> wrote: > On 29-07-2016 20:33, Jerry Sievers wrote: >> I've done several ~7TB pg_upgrades and with the hard link option and a >> framework that parallelizes the post-analyzer phase... >> >> ...45 minutes till completion. > GL to you Luck has nothing to do with anything. You really might want to post with more details and see whether people can help sort out why you have seen such slow performance where so many others have not. I hope it's not just a matter of saying "I have a 1TB database and upgrade is slow, therefore it is slow because it is 1TB." That would be roughly the equivalent of saying "I have a blue car and it is slow, therefore it is slow because it is blue." It just might be the flat tire that actually matters. If your upgrade is slow because you have 10 million database objects, that might be a hard one to overcome, but it might be something with an easy solution in the pg_upgrade options or server configuration. -- Kevin Grittner EDB: http://www.enterprisedb.com The Enterprise PostgreSQL Company -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
