##// END OF EJS Templates
feat(redis): added ability to specify a redis key prefix for all cache regions of dogpile that use redis backend
feat(redis): added ability to specify a redis key prefix for all cache regions of dogpile that use redis backend

File last commit:

r5493:18ecb0a9 default
r5582:15d02bff default
Show More
known-issues.rst
117 lines | 3.2 KiB | text/x-rst | RstLexer
project: added all source files and assets
r1 .. _known-issues:
Known Issues
============
Windows Upload
--------------
There can be an issue with uploading files from web interface on Windows,
and afterwards users cannot properly clone or synchronize with the repository.
Early testing shows that often uploading files via HTML forms on Windows
includes the full path of the file being uploaded and not the name of the file.
Old Format of Git Repositories
------------------------------
There is an issue when trying to import old |git| format |repos| into recent
versions of |RCE|. This issue can occur when importing from external |git|
repositories or from older versions of |RCE| (<=2.2.7).
To convert the old version into a current version, clone the old
|repo| into a local machine using a recent |git| client, then push it to a new
|repo| inside |RCE|.
docs: updated section about locale-archive problems
r2953
project: added all source files and assets
r1 VCS Server Memory Consumption
-----------------------------
The VCS Server cache grows without limits if not configured correctly. This
applies to |RCE| versions prior to the 3.3.2 releases, as 3.3.2
shipped with the optimal configuration as default. See the
:ref:`vcs-server-maintain` section for details.
To fix this issue, upgrade to |RCE| 3.3.2 or greater, and if you discover
memory consumption issues check the VCS Server settings.
docs: updated docs on local archives changes
r4295 Newer Operating system locales
------------------------------
docs: updated section about locale-archive problems
r2953
|RCC| has a know problem with locales, due to changes in glibc 2.27+ which affects
the local-archive format, which is now incompatible with our used glibc 2.26.
docs: updated docs on local archives changes
r4295 Mostly affected are:
Dmytro Sirant
- Added Kubernetes deployment section to the documentation...
r5493
docs: updated docs on local archives changes
r4295 - Fedora 23+
- Ubuntu 18.04
- CentOS / RHEL 8
project: added all source files and assets
r1
docs: updated section about locale-archive problems
r2953 To work around this problem, you need set path to ``$LOCAL_ARCHIVE`` to the
locale package in older pre glibc 2.27 format, or set `LC_ALL=C` in your enviroment.
To use the pre 2.27 locale-archive fix follow these steps:
project: added all source files and assets
r1
docs: updated section about locale-archive problems
r2953 1. Download the pre 2.27 locale-archive package
project: added all source files and assets
r1
docs: updated section about locale-archive problems
r2953 .. code-block:: bash
wget https://dls.rhodecode.com/assets/locale-archive
project: added all source files and assets
r1
2. Point ``$LOCAL_ARCHIVE`` to the locale package.
.. code-block:: bash
docs: updated section about locale-archive problems
r2953 $ export LOCALE_ARCHIVE=/home/USER/locale-archive # change to your path
docs: updated docs on local archives changes
r4295 This should be added *both* in `enviroment` variable of `~/.rccontrol/supervisor/supervisord.ini`
e.g
```
[supervisord]
environment = HOME=/home/user/rhodecode,LOCALE_ARCHIVE=/YOUR-PATH/locale-archive`
```
and in user .bashrc/.zshrc etc, or via a startup script that
docs: updated section about locale-archive problems
r2953 runs `rccontrol self-init`
project: added all source files and assets
r1
If you happen to be running |RCC| from systemd, use the following
example to pass the correct locale information on boot.
.. code-block:: ini
[Unit]
Description=Rhodecode
After=network.target
[Service]
Type=forking
User=scm
Environment="LOCALE_ARCHIVE=/YOUR-PATH/locale-archive"
ExecStart=/YOUR-PATH/.rccontrol-profile/bin/rccontrol-self-init
[Install]
WantedBy=multi-user.target
Dmytro Sirant
- Added Kubernetes deployment section to the documentation...
r5493
Merge stucks in "merging" status
--------------------------------
Similar issues:
- Pull Request duplicated and/or stucks in "creating" status.
Mostly affected are:
- Kubernetes AWS EKS setup with NFS as shared storage
- AWS EFS as shared storage
Workaround:
1. Manually clear the repo cache via UI:
:menuselection:`Repository Settings --> Caches --> Invalidate repository cache`
1. Open problematic PR and reset status to "created"
Now you can merge PR normally