aboutsummaryrefslogtreecommitdiff
path: root/cddl/contrib
diff options
context:
space:
mode:
authorAndriy Gapon <avg@FreeBSD.org>2019-10-16 07:02:02 +0000
committerAndriy Gapon <avg@FreeBSD.org>2019-10-16 07:02:02 +0000
commitf4902c2a07172325115df5f6dad0519d4a63244f (patch)
tree3a0c64e4975de2883418354520da0ff1aad28be3 /cddl/contrib
parent35b885b060938f1f618e83e93a871548e64005c4 (diff)
downloadsrc-f4902c2a07172325115df5f6dad0519d4a63244f.tar.gz
src-f4902c2a07172325115df5f6dad0519d4a63244f.zip
9691 fat zap should prefetch when iterating
illumos/illumos-gate@52abb70e073c2a88808c0d66fd810ba8c5080572 https://github.com/illumos/illumos-gate/commit/52abb70e073c2a88808c0d66fd810ba8c5080572 https://www.illumos.org/issues/9691 When iterating over a ZAP object, we're almost always certain to iterate over the entire object. If there are multiple leaf blocks, we can realize a performance win by issuing reads for all the leaf blocks in parallel when the iteration begins. For example, if we have 10,000 snapshots, "zfs destroy -nv pool/fs@1%9999" can take 30 minutes when the cache is cold. This change provides a >3x performance improvement, by issuing the reads for all ~64 blocks of each ZAP object in parallel. Author: Matthew Ahrens <mahrens@delphix.com>
Notes
Notes: svn path=/vendor-sys/illumos/dist/; revision=353619
Diffstat (limited to 'cddl/contrib')
0 files changed, 0 insertions, 0 deletions