[Bug 4692] New: transplant -s leads to excessive server-side bundling

[mercurial-bugs at selenic.com]

http://bz.selenic.com/show_bug.cgi?id=4692

          Priority: normal
            Bug ID: 4692
                CC: mercurial-devel at selenic.com
          Assignee: bugzilla at selenic.com
           Summary: transplant -s leads to excessive server-side bundling
          Severity: bug
    Classification: Unclassified
                OS: All
          Reporter: gregory.szorc at gmail.com
          Hardware: All
            Status: UNCONFIRMED
           Version: 3.4
         Component: transplant
           Product: Mercurial

This issue was originally reported at
https://bugzilla.mozilla.org/show_bug.cgi?id=1169669

STR:

1. hg clone https://hg.mozilla.org/integration/mozilla-inbound
2. cd mozilla-inbound
3. hg up 9ef73cc7ef3d
4. hg transplant -s https://reviewboard-hg.mozilla.org/gecko 8127d98a

The process "hangs" waiting for the server to process a "getbundle" command.

If I clone the gecko review repo (use --uncompressed since it is generaldelta)
and run a local server, I see the server try to generate a bundle with >6000
changesets. The client sends 90 x-hg-arg headers with common nodes to the
server. This is weird because `hg in -r 8127d98a
https://reviewboard-hg.mozilla.org/gecko` reports exactly 1 changeset:

changeset:   291358:8127d98a0f0c
tag:         tip
parent:      291267:0afa282390a3
user:        Gijs Kruitbosch <gijskruitbosch at gmail.com>
date:        Thu May 28 13:40:07 2015 +0100
summary:     Bug 1104732 - having deferred scripts shouldn't cause async
scripts to delay domcontentloaded, r?smaug

I reckon transplant or pull is doing something wonky with discovery.

-- 
You are receiving this mail because:
You are on the CC list for the bug.