lfs: fix the stall and corruption issue when concurrently uploading blobs
We've avoided the issue up to this point by gating worker usage with an
experimental config. See 10e62d5efa73, and the thread linked there for some of
the initial diagnosis, but essentially some data was being read from the blob
before an error occurred and `keepalive` retried, but didn't rewind the file
pointer. So the leading data was lost from the blob on the server, and the
connection stalled, trying to send more data than available.
In trying to recreate this, I was unable to do so uploading from Windows to
CentOS 7. But it reproduced every time going from CentOS 7 to another CentOS 7
over https.
I found recent fixes in the FaceBook repo to address this[1][2]. The commit
message for the first is:
The KeepAlive HTTP implementation is bugged in it's retry logic, it supports
reading from a file pointer, but doesn't support rewinding of the seek cursor
when it performs a retry. So it can happen that an upload fails for whatever
reason and will then 'hang' on the retry event.
The sequence of events that get triggered are:
- Upload file A, goes OK. Keep-Alive caches connection.
- Upload file B, fails due to (for example) failing Keep-Alive, but LFS file
pointer has been consumed for the upload and fd has been closed.
- Retry for file B starts, sets the Content-Length properly to the expected
file size, but since file pointer has been consumed no data will be uploaded,
causing the server to wait for the uploaded data until either client or
server reaches a timeout, making it seem as our mercurial process hangs.
This is just a stop-gap measure to prevent this behavior from blocking Mercurial
(LFS has retry logic). A proper solutions need to be build on top of this
stop-gap measure: for upload from file pointers, we should support fseek() on
the interface. Since we expect to consume the whole file always anyways, this
should be safe. This way we can seek back to the beginning on a retry.
I ported those two patches, and it works. But I see that `url._sendfile()` does
a rewind on `httpsendfile` objects[3], so maybe it's better to keep this all in
one place and avoid a second seek. We may still want the first FaceBook patch
as extra protection for this problem in general. The other two uses of
`httpsendfile` are in the wire protocol to upload bundles, and to upload
largefiles. Neither of these appear to use a worker, and I'm not sure why
workers seem to trigger this, or if this could have happened without a worker.
Since `httpsendfile` already has a `close()` method, that is dropped. That
class also explicitly says there's no `__len__` attribute, so that is removed
too. The override for `read()` is necessary to avoid the progressbar usage per
file.
[1] https://github.com/facebookexperimental/eden/commit/c350d6536d90c044c837abdd3675185644481469
[2] https://github.com/facebookexperimental/eden/commit/77f0d3fd0415e81b63e317e457af9c55c46103ee
[3] https://www.mercurial-scm.org/repo/hg/file/5.2.2/mercurial/url.py#l176
Differential Revision: https://phab.mercurial-scm.org/D7962
====================================
Testing head checking code: Case A-7
====================================
Mercurial checks for the introduction of new heads on push. Evolution comes
into play to detect if existing branches on the server are being replaced by
some of the new one we push.
This case is part of a series of tests checking this behavior.
Category A: simple case involving a branch being superceeded by another.
TestCase 7: multi-changeset branch, split on multiple other, (head on its own branch), same number of head
.. old-state:
..
.. * 2 branch (1-changeset, and 2-changesets)
..
.. new-state:
..
.. * 1 new branch superceeding the head of the old-2-changesets-branch,
.. * 1 new changesets on the old-1-changeset-branch superceeding the base of the other
..
.. expected-result:
..
.. * push allowed
..
.. graph-summary:
..
.. B ø⇠◔ B'
.. | |
.. A'◔⇢ø |
.. | |/
.. C ● |
.. \|
.. ●
$ . $TESTDIR/testlib/push-checkheads-util.sh
Test setup
----------
$ mkdir A7
$ cd A7
$ setuprepos
creating basic server and client repo
updating to branch default
2 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ cd server
$ mkcommit B0
$ hg up 0
0 files updated, 0 files merged, 2 files removed, 0 files unresolved
$ mkcommit C0
created new head
$ cd ../client
$ hg pull
pulling from $TESTTMP/A7/server
searching for changes
adding changesets
adding manifests
adding file changes
added 2 changesets with 2 changes to 2 files (+1 heads)
new changesets d73caddc5533:0f88766e02d6 (2 drafts)
(run 'hg heads' to see heads, 'hg merge' to merge)
$ hg up 'desc(C0)'
1 files updated, 0 files merged, 1 files removed, 0 files unresolved
$ mkcommit A1
$ hg up 0
0 files updated, 0 files merged, 2 files removed, 0 files unresolved
$ mkcommit B1
created new head
$ hg debugobsolete `getid "desc(A0)" ` `getid "desc(A1)"`
1 new obsolescence markers
obsoleted 1 changesets
1 new orphan changesets
$ hg debugobsolete `getid "desc(B0)" ` `getid "desc(B1)"`
1 new obsolescence markers
obsoleted 1 changesets
$ hg log -G --hidden
@ 25c56d33e4c4 (draft): B1
|
| o a0802eb7fc1b (draft): A1
| |
| o 0f88766e02d6 (draft): C0
|/
| x d73caddc5533 (draft): B0
| |
| x 8aaa48160adc (draft): A0
|/
o 1e4be0697311 (public): root
Actual testing
--------------
$ hg push
pushing to $TESTTMP/A7/server
searching for changes
adding changesets
adding manifests
adding file changes
added 2 changesets with 2 changes to 2 files (+1 heads)
2 new obsolescence markers
obsoleted 2 changesets
$ cd ../..