9.2 KiB
stage | group | info | type |
---|---|---|---|
Create | Source Code | To determine the technical writer assigned to the Stage/Group associated with this page, see https://about.gitlab.com/handbook/product/ux/technical-writing/#assignments | howto |
Troubleshooting Git (FREE)
Sometimes things don't work the way they should or as you might expect when you're using Git. Here are some tips on troubleshooting and resolving issues with Git.
Broken pipe errors on git push
'Broken pipe' errors can occur when attempting to push to a remote repository. When pushing you usually see:
Write failed: Broken pipe
fatal: The remote end hung up unexpectedly
To fix this issue, here are some possible solutions.
Increase the POST buffer size in Git
If you're using Git over HTTP instead of SSH, you can try increasing the POST buffer size in Git's configuration.
Example of an error during a clone:
fatal: pack has bad object at offset XXXXXXXXX: inflate returned -5
Open a terminal and enter:
git config http.postBuffer 52428800
The value is specified in bytes, so in the above case the buffer size has been set to 50MB. The default is 1MB.
RPC failed; curl 92 HTTP/2 stream 0 was not closed cleanly: INTERNAL_ERROR (err 2)
This problem may be caused by a slow internet connection. If you use Git over HTTP instead of SSH, try one of these fixes:
- Increase the POST buffer size in the Git configuration with
git config http.postBuffer 52428800
. - Switch to the
HTTP/1.1
protocol withgit config http.version HTTP/1.1
.
If neither approach fixes the error, you may need a different internet service provider.
Check your SSH configuration
If pushing over SSH, first check your SSH configuration as 'Broken pipe' errors can sometimes be caused by underlying issues with SSH (such as authentication). Make sure that SSH is correctly configured by following the instructions in the SSH troubleshooting documentation.
If you're a GitLab administrator with server access, you can also prevent
session timeouts by configuring SSH keep-alive
on the client or the server.
NOTE: Configuring both the client and the server is unnecessary.
To configure SSH on the client side:
-
On UNIX, edit
~/.ssh/config
(create the file if it doesn't exist) and add or edit:Host your-gitlab-instance-url.com ServerAliveInterval 60 ServerAliveCountMax 5
-
On Windows, if you are using PuTTY, go to your session properties, then navigate to "Connection" and under "Sending of null packets to keep session active", set
Seconds between keepalives (0 to turn off)
to60
.
To configure SSH on the server side, edit /etc/ssh/sshd_config
and add:
ClientAliveInterval 60
ClientAliveCountMax 5
Running a git repack
If 'pack-objects' type errors are also being displayed, you can try to
run a git repack
before attempting to push to the remote repository again:
git repack
git push
Upgrade your Git client
In case you're running an older version of Git (< 2.9), consider upgrading to >= 2.9 (see Broken pipe when pushing to Git repository).
ssh_exchange_identification
error
Users may experience the following error when attempting to push or pull using Git over SSH:
Please make sure you have the correct access rights
and the repository exists.
...
ssh_exchange_identification: read: Connection reset by peer
fatal: Could not read from remote repository.
or
ssh_exchange_identification: Connection closed by remote host
fatal: The remote end hung up unexpectedly
or
kex_exchange_identification: Connection closed by remote host
Connection closed by x.x.x.x port 22
This error usually indicates that SSH daemon's MaxStartups
value is throttling
SSH connections. This setting specifies the maximum number of concurrent, unauthenticated
connections to the SSH daemon. This affects users with proper authentication
credentials (SSH keys) because every connection is 'unauthenticated' in the
beginning. The default value is 10
.
Increase MaxStartups
on the GitLab server
by adding or modifying the value in /etc/ssh/sshd_config
:
MaxStartups 100:30:200
100:30:200
means up to 100 SSH sessions are allowed without restriction,
after which 30% of connections are dropped until reaching an absolute maximum of 200.
After you modify the value of MaxStartups
, check for any errors in the configuration.
sudo sshd -t -f /etc/ssh/sshd_config
If the configuration check runs without errors, it should be safe to restart the SSH daemon for the change to take effect.
# Debian/Ubuntu
sudo systemctl restart ssh
# CentOS/RHEL
sudo service sshd restart
Timeout during git push
/ git pull
If pulling/pushing from/to your repository ends up taking more than 50 seconds, a timeout is issued. It contains a log of the number of operations performed and their respective timings, like the example below:
remote: Running checks for branch: master
remote: Scanning for LFS objects... (153ms)
remote: Calculating new repository size... (cancelled after 729ms)
This could be used to further investigate what operation is performing poorly and provide GitLab with more information on how to improve the service.
git clone
over HTTP fails with transfer closed with outstanding read data remaining
error
Sometimes, when cloning old or large repositories, the following error is thrown:
error: RPC failed; curl 18 transfer closed with outstanding read data remaining
fatal: The remote end hung up unexpectedly
fatal: early EOF
fatal: index-pack failed
This is a common problem with Git itself, due to its inability to handle large files or large quantities of files. Git LFS was created to work around this problem; however, even it has limitations. It's usually due to one of these reasons:
- The number of files in the repository.
- The number of revisions in the history.
- The existence of large files in the repository.
The root causes vary, so multiple potential solutions exist, and you may need to apply more than one:
-
If this error occurs when cloning a large repository, you can decrease the cloning depth to a value of
1
. For example:variables: GIT_DEPTH: 1
-
You can increase the http.postBuffer value in your local Git configuration from the default 1 MB value to a value greater than the repository size. For example, if
git clone
fails when cloning a 500 MB repository, you should sethttp.postBuffer
to524288000
:# Set the http.postBuffer size, in bytes git config http.postBuffer 524288000
-
You can increase the
http.postBuffer
on the server side:-
Modify the GitLab instance's
gitlab.rb
file:gitaly['gitconfig'] = [ # Set the http.postBuffer size, in bytes {key: "http.postBuffer", value: "524288000"}, ]
-
After applying this change, apply the configuration change:
sudo gitlab-ctl reconfigure
-
For example, if a repository has a very long history and no large files, changing
the depth should fix the problem. However, if a repository has very large files,
even a depth of 1 may be too large, thus requiring the postBuffer
change.
If you increase your local postBuffer
but the NGINX value on the backend is still
too small, the error persists.
Modifying the server is not always an option, and introduces more potential risk. Attempt local changes first.
Password expired error on Git fetch via SSH for LDAP user
If git fetch
returns this HTTP 403 Forbidden
error on a self-managed instance of
GitLab, the password expiration date (users.password_expires_at
) for this user in the
GitLab database is a date in the past:
Your password expired. Please access GitLab from a web browser to update your password.
Requests made with a SSO account and where password_expires_at
is not null
return this error:
"403 Forbidden - Your password expired. Please access GitLab from a web browser to update your password."
To resolve this issue, you can update the password expiration by either:
-
Using the
gitlab-rails console
:gitlab-rails console user.update!(password_expires_at: nil)
-
Using
gitlab-psql
:# gitlab-psql UPDATE users SET password_expires_at = null WHERE username='<USERNAME>';
The bug was reported in this issue.
Error on Git fetch: "HTTP Basic: Access Denied"
If you receive an HTTP Basic: Access denied
error when using Git over HTTP(S),
refer to the two-factor authentication troubleshooting guide.