Mercurial > hg-stable
changeset 44085:0ee0a3f6a990
lfs: check content length after downloading content
Adapted from the Facebook repo[1]. The intent is to distinguish between the
connection dying and getting served a corrupt blob.
The original message:
HTTP makes no provision to tell your client that you failed halfway through
producing your response and won't have the answer they're looking for. So, if a
LFS server fails while producing a response, then we'll report an OID mismatch.
We can do a little better and disambiguate between "the server sent us the
wrong blob" (very scary) and "the server crashed" (merely annoying) by looking
at the content length of the response we got back. If it's not what was
advertised, we can reasonably safely assume the server crashed.
[1] https://github.com/facebookexperimental/eden/commit/2a4a6fab4e882ed89b948bfc1e7d56d7c3c99dd2
Differential Revision: https://phab.mercurial-scm.org/D7881
author | Matt Harbison <matt_harbison@yahoo.com> |
---|---|
date | Tue, 14 Jan 2020 19:42:24 -0500 |
parents | 84f2becbd106 |
children | ffac09da7a19 |
files | hgext/lfs/blobstore.py hgext/lfs/wireprotolfsserver.py tests/test-lfs-serve-access.t |
diffstat | 3 files changed, 23 insertions(+), 8 deletions(-) [+] |
line wrap: on
line diff
--- a/hgext/lfs/blobstore.py Tue Jan 14 18:02:20 2020 -0500 +++ b/hgext/lfs/blobstore.py Tue Jan 14 19:42:24 2020 -0500 @@ -155,15 +155,29 @@ return self.vfs(oid, b'rb') - def download(self, oid, src): + def download(self, oid, src, content_length): """Read the blob from the remote source in chunks, verify the content, and write to this local blobstore.""" sha256 = hashlib.sha256() + size = 0 with self.vfs(oid, b'wb', atomictemp=True) as fp: for chunk in util.filechunkiter(src, size=1048576): fp.write(chunk) sha256.update(chunk) + size += len(chunk) + + # If the server advertised a length longer than what we actually + # received, then we should expect that the server crashed while + # producing the response (but the server has no way of telling us + # that), and we really don't need to try to write the response to + # the localstore, because it's not going to match the expected. + if content_length is not None and int(content_length) != size: + msg = ( + b"Response length (%s) does not match Content-Length " + b"header (%d): likely server-side crash" + ) + raise LfsRemoteError(_(msg) % (size, int(content_length))) realoid = node.hex(sha256.digest()) if realoid != oid: @@ -492,6 +506,7 @@ response = b'' try: with contextlib.closing(self.urlopener.open(request)) as res: + contentlength = res.info().get(b"content-length") ui = self.ui # Shorten debug lines if self.ui.debugflag: ui.debug(b'Status: %d\n' % res.status) @@ -503,7 +518,7 @@ if action == b'download': # If downloading blobs, store downloaded data to local # blobstore - localstore.download(oid, res) + localstore.download(oid, res, contentlength) else: while True: data = res.read(1048576) @@ -637,7 +652,7 @@ def readbatch(self, pointers, tostore): for p in _deduplicate(pointers): with self.vfs(p.oid(), b'rb') as fp: - tostore.download(p.oid(), fp) + tostore.download(p.oid(), fp, None) class _nullremote(object):
--- a/hgext/lfs/wireprotolfsserver.py Tue Jan 14 18:02:20 2020 -0500 +++ b/hgext/lfs/wireprotolfsserver.py Tue Jan 14 19:42:24 2020 -0500 @@ -327,7 +327,7 @@ statusmessage = hgwebcommon.statusmessage try: - localstore.download(oid, req.bodyfh) + localstore.download(oid, req.bodyfh, req.headers[b'Content-Length']) res.status = statusmessage(HTTP_OK if existed else HTTP_CREATED) except blobstore.LfsCorruptionError: _logexception(req)
--- a/tests/test-lfs-serve-access.t Tue Jan 14 18:02:20 2020 -0500 +++ b/tests/test-lfs-serve-access.t Tue Jan 14 19:42:24 2020 -0500 @@ -210,7 +210,7 @@ > > store = repo.svfs.lfslocalblobstore > class badstore(store.__class__): - > def download(self, oid, src): + > def download(self, oid, src, contentlength): > '''Called in the server to handle reading from the client in a > PUT request.''' > origread = src.read @@ -218,7 +218,7 @@ > # Simulate bad data/checksum failure from the client > return b'0' * len(origread(nbytes)) > src.read = _badread - > super(badstore, self).download(oid, src) + > super(badstore, self).download(oid, src, contentlength) > > def _read(self, vfs, oid, verify): > '''Called in the server to read data for a GET request, and then @@ -351,8 +351,8 @@ $LOCALIP - - [$ERRDATE$] HG error: (glob) $LOCALIP - - [$ERRDATE$] HG error: Exception happened while processing request '/.hg/lfs/objects/b5bb9d8014a0f9b1d61e21e796d78dccdf1352f23cd32812f4850b878ae4944c': (glob) $LOCALIP - - [$ERRDATE$] HG error: Traceback (most recent call last): (glob) - $LOCALIP - - [$ERRDATE$] HG error: localstore.download(oid, req.bodyfh) (glob) - $LOCALIP - - [$ERRDATE$] HG error: super(badstore, self).download(oid, src) (glob) + $LOCALIP - - [$ERRDATE$] HG error: localstore.download(oid, req.bodyfh, req.headers[b'Content-Length']) + $LOCALIP - - [$ERRDATE$] HG error: super(badstore, self).download(oid, src, contentlength) $LOCALIP - - [$ERRDATE$] HG error: _(b'corrupt remote lfs object: %s') % oid (glob) $LOCALIP - - [$ERRDATE$] HG error: LfsCorruptionError: corrupt remote lfs object: b5bb9d8014a0f9b1d61e21e796d78dccdf1352f23cd32812f4850b878ae4944c (no-py3 !) $LOCALIP - - [$ERRDATE$] HG error: hgext.lfs.blobstore.LfsCorruptionError: corrupt remote lfs object: b5bb9d8014a0f9b1d61e21e796d78dccdf1352f23cd32812f4850b878ae4944c (py3 !)