This is the mail archive of the
mailing list for the Cygwin project.
Re: "git clone" over ssh fails intermittently w/snapshot 20120123
Christopher Faylor wrote:
>> On Fri, Jan 27, 2012 at 02:32:54PM -0800, Kevin Layer wrote:
>> >I don't know what happened to the email in transit, but much
>> >information at the head of the email was chopped off. From my outbox,
>> >here's the missing bit:
>> >This is the failure mode:
>> >$ rm -fr test; git clone git:/repo/git/composer test
>> >Cloning into test...
>> >fatal: The remote end hung up unexpectedly
>> >fatal: early EOF
>> >fatal: index-pack failed
>> >That is not using the git protocol. The machine is called `git'.
>> >It fails most of the time, but it definitely works sometimes, and even
>> >has stretches of successes.
>> >At first, I thought it was isolated to large repos, but it definitely
>> >happened on some smaller ones.
>> You really haven't provided any details beyond "git clone doesn't work"
>> along with cygcheck output.
Well, I've shown the error message, and now answer the following
questions. Hopefully that's enough.
>> You've previously reported that git worked great. What changed?
This is a different machine.
>> What is the server that you're cloning from? Another windows machine
>> running Cygwin? A linux box? Solaris?
The server is a linux box.
>> What does "large repo" mean to you? I just tried this on the gcc repo
>> which is, from my point of view, pretty big. The clone succeeded without
As I said, it does *not* have to do with size.
>> Where is the server located? Is it on a local network or is it from
>> somewhere on the internet?
Local area network.
>> From my reading of the git documentation, using the host:/path syntax
>> means you're using ssh to connect to the other machine. Have you
>> checked the logs of the server to see if anything is reported?
Have now and nothing.
>> Is this new behavior just seen with this snapshot or is it something you
>> noticed in 1.7.9? If you've just seen it in the latest snapshot it
>> would be useful (as I've previously pleaded) to know in which snapshot
>> the bad behavior first showed up.
Never seen until the snapshot was installed.
Before the failure, I had removed and reinstalled cygwin, then
installed the snapshot. The machine on which git works was upgraded
little at a time.
Problem reports: http://cygwin.com/problems.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple