Jehan-Guillaume (ioguix) de Rorthais
Jehan-Guillaume (ioguix) de Rorthais
Hi, First, in the same fashion than top, you could add some interactive commands to enable/disable some infos and save some space. These keys can also turn to command argument...
Hi, As I understand the discussion on pgsql-hackers, this patch hasn't been back-patched to v10. So it's OK for v11 and after.
Hi @shangzixie, It comes from `pg_stat_all_tables.n_live_tup`. > after I REINDEX an index, the output bloat_size about the index doesn't change. On what version are you testing? Could you share your...
Hi, I might be wrong, but it seems related to #9. I don't understand why you compare with the 8.2-8.3 query as it seem your problem is on a recent...
Hi, for now, I added some explanation in the README. > Perhaps is it possible to exclude text columns with 18.0 bytes and an extended/external storage ? Any idea and...
Hi @shangzixie, Right, it seems `t_tid` is missing here. According to the source code, it adds 3 bytes to the tuple header: ~~~C // From src/include/storage/itemptr.h /* * ItemPointer: *...
> I don't know if you could add 6 bytes rather than 5 bytes. Maybe the t_tid is 6 bytes? Pfff, obviously, my bad, I've been heedless... Worth thing is...
I keep this issue open, I'll close it once fixed. Thank you!
Hi Marko, Any feedback about this patch ?
Hi, I don't need the avg\* stats, I was just following how stats already exists in pgbouncer. I don't mind dropping the avg stats and add a delta_time field or...