> 19 июля 2018 г., в 16:28, Heikki Linnakangas <hlinn...@iki.fi> написал(а):
> Hmm. So, while we are scanning the right sibling, which was moved to
> lower-numbered block because of a concurrent split, the original page is
> split again? That's OK, we've already scanned all the tuples on the original
> page, before we recurse to deal with the right sibling. (The corresponding
> B-tree code also releases the lock on the original page when recursing)
Seems right.
>
> I did some refactoring, to bring this closer to the B-tree code, for the sake
> of consistency. See attached patch. This also eliminates the 2nd pass by
> gistvacuumcleanup(), in case we did that in the bulkdelete-phase already.
Thanks!
>
> There was one crucial thing missing: in the outer loop, we must ensure that
> we scan all pages, even those that were added after the vacuum started.
Correct. Quite a neat logic behind the order of acquiring npages, comparing and
vacuuming page. Notes in FIXME look correct except function names.
> There's a comment explaining that in btvacuumscan(). This version fixes that.
>
> I haven't done any testing on this. Do you have any test scripts you could
> share?
I use just a simple tests that setup replication and does random inserts and
vaccums. Not a rocket science, just a mutated script
for i in $(seq 1 12); do
size=$((100 * 2**$i))
./psql postgres -c "create table x as select cube(random()) c from
generate_series(1,$size) y; create index on x using gist(c);"
./psql postgres -c "delete from x;"
./psql postgres -c "VACUUM x;"
./psql postgres -c "VACUUM x;"
./psql postgres -c "drop table x;"
./psql postgres -c "create table x as select cube(random()) c from
generate_series(1,$size) y; create index on x using gist(c);"
./psql postgres -c "delete from x where (c~>1)>0.1;"
./psql postgres -c "VACUUM x;"
./psql postgres -c "insert into x select cube(random()) c from
generate_series(1,$size) y;"
./psql postgres -c "VACUUM x;"
./psql postgres -c "delete from x where (c~>1)>0.1;"
./psql postgres -c "select pg_size_pretty(pg_relation_size('x_c_idx'));"
./psql postgres -c "VACUUM FULL x;"
./psql postgres -c "select pg_size_pretty(pg_relation_size('x_c_idx'));"
./psql postgres -c "drop table x;"
done
> I think we need some repeatable tests for the concurrent split cases.
It is hard to trigger left splits until we delete pages. I'll try to hack
gistNewBuffer() to cause something similar.
> Even if it involves gdb or some other hacks that we can't include in the
> regression test suite, we need something now, while we're hacking on this.
>
> One subtle point, that I think is OK, but gave me a pause, and probably
> deserves comment somewhere: A concurrent root split can turn a leaf page into
> one internal (root) page, and two new leaf pages. The new root page is placed
> in the same block as the old page, while both new leaf pages go to freshly
> allocated blocks. If that happens while vacuum is running, might we miss the
> new leaf pages? As the code stands, we don't do the "follow-right" dance on
> internal pages, so we would not recurse into the new leaf pages. At first, I
> thought that's a problem, but I think we can get away with it. The only
> scenario where a root split happens on a leaf page, is when the index has
> exactly one page, a single leaf page. Any subsequent root splits will split
> an internal page rather than a leaf page, and we're not bothered by those. In
> the case that a root split happens on a single-page index, we're OK, because
> we will always scan that page either before, or after the split. If we scan
> the single page before the split, we see all the leaf tuples on that page. If
> we scan the single page after the split, it means that we start the scan
> after the split, and we will see both leaf pages as we continue the scan.
Yes, only page 0 may change type, and page 0 cannot split to left.
I'm working on triggering left split during vacuum. Will get back when done.
Thanks!
Best regards, Andrey Borodin.
