This is an archive of the discontinued Mercurial Phabricator instance.

Differential D4976

exchangev2: honor server advertised manifestdata recommended batch size
ClosedPublic

Authored by indygreg on Oct 12 2018, 4:14 AM.

Details

Reviewers
None
Group Reviewers
hg-reviewers
Commits
rHGb797150a1ab9: exchangev2: honor server advertised manifestdata recommended batch size
Summary

Let's plug the client up to the server-advertised recommended batch size
for manifestdata requests.

Diff Detail

Repository
rHG Mercurial
Lint
Automatic diff as part of commit; lint not applicable.
Unit
Automatic diff as part of commit; unit tests not applicable.

Revision Contents
Changeset List

PathPackages
Mmercurial/exchangev2.py (5 lines)

Diff 11883

mercurial/exchangev2.py

0 0
) )
progress.increment() progress.increment()
progress = repo.ui.makeprogress(_('manifests'), unit=_('chunks'), progress = repo.ui.makeprogress(_('manifests'), unit=_('chunks'),
total=len(fetchnodes)) total=len(fetchnodes))
# Fetch manifests 10,000 per command. commandmeta = remote.apidescriptor[b'commands'][b'manifestdata']
# TODO have server advertise preferences? batchsize = commandmeta.get(b'recommendedbatchsize', 10000)
# TODO make size configurable on client? # TODO make size configurable on client?
batchsize = 10000
# We send commands 1 at a time to the remote. This is not the most # We send commands 1 at a time to the remote. This is not the most
# efficient because we incur a round trip at the end of each batch. # efficient because we incur a round trip at the end of each batch.
# However, the existing frame-based reactor keeps consuming server # However, the existing frame-based reactor keeps consuming server
# data in the background. And this results in response data buffering # data in the background. And this results in response data buffering
# in memory. This can consume gigabytes of memory. # in memory. This can consume gigabytes of memory.
# TODO send multiple commands in a request once background buffering # TODO send multiple commands in a request once background buffering
# issues are resolved. # issues are resolved.