##// END OF EJS Templates
httppeer: remove support for connecting to <0.9.1 servers (BC)...
httppeer: remove support for connecting to <0.9.1 servers (BC) Previously, HTTP wire protocol clients would attempt a "capabilities" wire protocol command. If that failed, they would fall back to issuing a "between" command. The "capabilities" command was added in Mercurial 0.9.1 (released July 2006). The "between" command has been present for as long as the wire protocol has existed. So if the "between" command failed, it was safe to assume that the remote could not speak any version of the Mercurial wire protocol. The "between" fallback was added in 395a84f78736 in 2011. Before that changeset, Mercurial would *always* issue the "between" command and would issue "capabilities" if capabilities were requested. At that time, many connections would issue "capabilities" eventually, so it was decided to issue "capabilities" by default and fall back to "between" if that failed. This saved a round trip when connecting to modern servers while still preserving compatibility with legacy servers. Fast forward ~7 years. Mercurial servers supporting "capabilities" have been around for over a decade. If modern clients are connecting to <0.9.1 servers, they are getting a bad experience. They may even be getting bad data (an old server is vulnerable to numerous security issues and could have been p0wned, leading to a Mercurial repository serving backdoors or other badness). In addition, the fallback can harm experience for modern servers. If a client experiences an intermittent HTTP request failure (due to bad network, etc) and falls back to a "between" that works, it would assume an empty capability set and would attempt to communicate with the repository using a very ancient wire protocol. Auditing HTTP logs for hg.mozilla.org, I did find a handful of requests for the null range of the "between" command. However, requests can be days apart. And when I do see requests, they come in batches. Those batches seem to correlate to spikes of HTTP 500 or other server/network events. So I think these requests are fallbacks from failed "capabilities" requests and not from old clients. If you need even more evidence to discontinue support, apparently we have no test coverage for communicating with servers not supporting "capabilities." I know this because all tests pass with the "between" fallback removed. Finally, server-side support for <0.9.1 pushing (the "addchangegroup" wire protocol command along with locking-related commands) was dropped from the HTTP client in fda0867cfe03 in 2017 and the SSH client in 9f6e0e7ef828 in 2015. I think this all adds up to enough justification for removing client support for communicating with servers not supporting "capabilities." So this commit removes that fallback. Differential Revision: https://phab.mercurial-scm.org/D2001

File last commit:

r26587:56b2bcea default
r35902:197d10e1 default
Show More
treediscovery.py
159 lines | 5.2 KiB | text/x-python | PythonLexer
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164 # discovery.py - protocol changeset discovery functions
#
# Copyright 2010 Matt Mackall <mpm@selenic.com>
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.
Gregory Szorc
treediscovery: use absolute_import
r25987 from __future__ import absolute_import
Martin von Zweigbergk
util: drop alias for collections.deque...
r25113 import collections
Gregory Szorc
treediscovery: use absolute_import
r25987
from .i18n import _
from .node import (
nullid,
short,
)
from . import (
error,
)
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164
def findcommonincoming(repo, remote, heads=None, force=False):
"""Return a tuple (common, fetch, heads) used to identify the common
subset of nodes between repo and remote.
"common" is a list of (at least) the heads of the common subset.
"fetch" is a list of roots of the nodes that would be incoming, to be
supplied to changegroupsubset.
"heads" is either the supplied heads, or else the remote's heads.
"""
Pierre-Yves David
discovery: stop using nodemap for membership testing...
r20225 knownnode = repo.changelog.hasnode
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164 search = []
fetch = set()
seen = set()
seenbranch = set()
base = set()
if not heads:
heads = remote.heads()
if repo.changelog.tip() == nullid:
base.add(nullid)
if heads != [nullid]:
return [nullid], [nullid], list(heads)
Peter Arrenbrecht
treediscovery: fix regression when run against older repos (issue2793)...
r14199 return [nullid], [], heads
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164
# assume we're closer to the tip than the root
# and start by examining the heads
repo.ui.status(_("searching for changes\n"))
unknown = []
for h in heads:
Pierre-Yves David
discovery: stop using nodemap for membership testing...
r20225 if not knownnode(h):
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164 unknown.append(h)
else:
base.add(h)
Peter Arrenbrecht
treediscovery: fix regression when run against older repos (issue2793)...
r14199 if not unknown:
return list(base), [], list(heads)
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164 req = set(unknown)
reqcnt = 0
# search through remote branches
# a 'branch' here is a linear segment of history, with four parts:
# head, root, first parent, second parent
# (a branch always has two parents (or none) by definition)
Martin von Zweigbergk
util: drop alias for collections.deque...
r25113 unknown = collections.deque(remote.branches(unknown))
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164 while unknown:
r = []
while unknown:
Bryan O'Sullivan
cleanup: use the deque type where appropriate...
r16803 n = unknown.popleft()
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164 if n[0] in seen:
continue
repo.ui.debug("examining %s:%s\n"
% (short(n[0]), short(n[1])))
if n[0] == nullid: # found the end of the branch
pass
elif n in seenbranch:
repo.ui.debug("branch already found\n")
continue
Pierre-Yves David
discovery: stop using nodemap for membership testing...
r20225 elif n[1] and knownnode(n[1]): # do we know the base?
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164 repo.ui.debug("found incomplete branch %s:%s\n"
% (short(n[0]), short(n[1])))
search.append(n[0:2]) # schedule branch range for scanning
seenbranch.add(n)
else:
if n[1] not in seen and n[1] not in fetch:
Pierre-Yves David
discovery: stop using nodemap for membership testing...
r20225 if knownnode(n[2]) and knownnode(n[3]):
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164 repo.ui.debug("found new changeset %s\n" %
short(n[1]))
fetch.add(n[1]) # earliest unknown
for p in n[2:4]:
Pierre-Yves David
discovery: stop using nodemap for membership testing...
r20225 if knownnode(p):
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164 base.add(p) # latest known
for p in n[2:4]:
Pierre-Yves David
discovery: stop using nodemap for membership testing...
r20225 if p not in req and not knownnode(p):
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164 r.append(p)
req.add(p)
seen.add(n[0])
if r:
reqcnt += 1
repo.ui.progress(_('searching'), reqcnt, unit=_('queries'))
repo.ui.debug("request %d: %s\n" %
(reqcnt, " ".join(map(short, r))))
for p in xrange(0, len(r), 10):
for b in remote.branches(r[p:p + 10]):
repo.ui.debug("received %s:%s\n" %
(short(b[0]), short(b[1])))
unknown.append(b)
# do binary search on the branches we found
while search:
newsearch = []
reqcnt += 1
repo.ui.progress(_('searching'), reqcnt, unit=_('queries'))
for n, l in zip(search, remote.between(search)):
l.append(n[1])
p = n[0]
f = 1
for i in l:
repo.ui.debug("narrowing %d:%d %s\n" % (f, len(l), short(i)))
Pierre-Yves David
discovery: stop using nodemap for membership testing...
r20225 if knownnode(i):
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164 if f <= 2:
repo.ui.debug("found new branch changeset %s\n" %
short(p))
fetch.add(p)
base.add(i)
else:
repo.ui.debug("narrowed branch search to %s:%s\n"
% (short(p), short(i)))
newsearch.append((p, i))
break
p, f = i, f * 2
search = newsearch
# sanity check our fetch list
for f in fetch:
Pierre-Yves David
discovery: stop using nodemap for membership testing...
r20225 if knownnode(f):
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164 raise error.RepoError(_("already have changeset ")
+ short(f[:4]))
base = list(base)
if base == [nullid]:
if force:
repo.ui.warn(_("warning: repository is unrelated\n"))
else:
Pierre-Yves David
error: get Abort from 'error' instead of 'util'...
r26587 raise error.Abort(_("repository is unrelated"))
Peter Arrenbrecht
discovery: add new set-based discovery...
r14164
repo.ui.debug("found new changesets starting at " +
" ".join([short(f) for f in fetch]) + "\n")
repo.ui.progress(_('searching'), None)
repo.ui.debug("%d total queries\n" % reqcnt)
return base, list(fetch), heads