-
Notifications
You must be signed in to change notification settings - Fork 1.8k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Documentation and examples update for use with docker compose v2 #2201
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Why did you add the logging
config everywhere?
.examples/docker-compose/with-nginx-proxy/postgres/apache/compose.yaml
Outdated
Show resolved
Hide resolved
.examples/docker-compose/with-nginx-proxy/mariadb/apache/compose.yaml
Outdated
Show resolved
Hide resolved
.examples/docker-compose/with-nginx-proxy/postgres/apache/compose.yaml
Outdated
Show resolved
Hide resolved
IMHO it's bad to duplicate all default
should be enough.
We use whatever Nextcloud recommends in their documenation. We also don't pin minor releases because then somebody would need to take care of updates.
Fine for me but I'm not sure how used people are to
Initially it wasn't added everywhere because it's optional and to keep the examples simple. But technically a database server is also optional.
Logging is out of scope for the examples. We shouldn't make any assumptions about their logging system. |
You are quite correct here, this is the most "clean" way of doing it. Instead of adding js and mjs, only mjs is added. Thanks !
I thought so, reverted back to 10.6 as per docs.
Docker put the notice of deprecation of V1 syntax quite a long time ago. As per docker docs and requirements, it must be named
Removed the logging sections too. While I agree with the fact we should not make assumptions, and people using these should take care of that, I think we might mention this problems with the logs somewhere. On a busy instances with more than few users these logs can get quite large and actually cause problems. Nginx and LE compaion tend to output a lot of stuff, most of the time harmless, but they can pile up if all goes well for longer time periods. Also fixed the random typos here and there, didn't noticed them the last 3 times I've read the changes. EDIT: I removed the conf volume. It's also based on some assumptions that shouldn't be assumed. It is usually helpful if someone is injecting additional stuff in the proxy, but in a general-use-case where the proxy is used only for NC, that volume is not needed. |
.examples/docker-compose/with-nginx-proxy/mariadb/apache/compose.yaml
Outdated
Show resolved
Hide resolved
Needs to be reverted in the README too |
I still have the mjs problem and spent 4 hours yesterday (just to say I'm not just lazily expecting help) to see how modifying nginx.conf might help with the well known error message about the mjs mime type. No success. I only mount the nginx.conf specifically in the container from outside, not the mime.types. I'm working off of this one with the fpm alpine docker compose: .examples/docker-compose/with-nginx-proxy/mariadb/fpm/web/nginx.conf Some of the posts I read through suggest it has something to do with DNS / domain entries, but a curl to an mjs file from inside the container yields results, so it doesn't seem it's a URL issue. |
Use the |
Reverted the README me, my bad. Also removed the GH-specific markdown. Fixed the little ":" too. |
Bump, i guess ? |
…ements. Signed-off-by: Kaloyan Nikolov <[email protected]>
Signed-off-by: Kaloyan Nikolov <[email protected]>
Signed-off-by: Kaloyan Nikolov <[email protected]>
Signed-off-by: Kaloyan Nikolov <[email protected]>
Signed-off-by: Kaloyan Nikolov <[email protected]>
Signed-off-by: Kaloyan Nikolov <[email protected]>
Signed-off-by: Kaloyan Nikolov <[email protected]>
Signed-off-by: Kaloyan Nikolov <[email protected]>
Signed-off-by: Kaloyan Nikolov <[email protected]>
Signed-off-by: Kaloyan Nikolov <[email protected]>
…ed version Signed-off-by: Kaloyan Nikolov <[email protected]>
Signed-off-by: Kaloyan Nikolov <[email protected]>
…ose.yaml Signed-off-by: J0WI <[email protected]>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Please revert --innodb-read-only-compressed=OFF
too
I added them, but I've also noticed the PR for MariaDB 10.11. Should i just change to 10.11 here as well ? @J0WI EDIT: I'm confused now. What is the correct mysql command, Edit 2 : matched to what it's in the current docs. It's been a while since I gave up on MariaDB and moved all the prod instances to postgresql. Still confused on the |
Signed-off-by: Kaloyan Nikolov <[email protected]>
The command should match what's documented in https://docs.nextcloud.com/server/latest/admin_manual/configuration_database/linux_database_configuration.html#configuring-a-mysql-or-mariadb-database. |
Signed-off-by: Kaloyan Nikolov <[email protected]>
Signed-off-by: Kaloyan Nikolov <[email protected]>
Updated it to match to the docs link you provided. In this repo version (front page) there is this section where it says Also sorry, didn't noticed i pushed the wrong change last time, totally my fault. Will be a bit more careful from now on. |
They are re-added in the README |
Signed-off-by: Kaloyan Nikolov <[email protected]>
Found it #1881
|
Signed-off-by: Kaloyan Nikolov <[email protected]>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If you add --log-bin
back it should keep the name binlog
Signed-off-by: Kaloyan Nikolov <[email protected]>
I think this is ready now :) |
Thanks! |
Howdy,
I noticed #2194 and the PR but there are more problems with the examples than just the
version
parameter. Docker changed the file name requirements too, changed the command and some of the syntax, jwilder's proxy is now part of nginx-proxy and more. This PR addresses those changes, and few others as well.What was changed:
Docs: ( README.MD ) have commands, file names and examples updated everywhere. Minor edits to match the examples section and links.
Files: All
docker-compose.yml
files are renamed tocompose.yaml
files as per docker requirements.Nginx configuration: I've been following the nginx mailing list since the problems we had with
mjs
file extension last year, and they wont fix it despite browsers expecting it astext/javascript
. I have modified the examples to include mime-types from Nginx to the web container, and removed the "workaround" from thenginx.conf
file to keep it as consistent as possible with the official Nextcloud docs. Effectively there is no difference from end user point of view.MariaDB: - Over the years we have several instances running using MariaDB and experience shows that 10.8.2 works better than 10.6 with recent versions of Nextcloud. I fully understand that it's not LTS release but we had several random problems with 10.6 that caused larger downtimes and then we switched to 10.8.2 in production, no issues whatsoever since then. It is also intentional that I didn't go newer than 10.8.2 because I cannot "promote" something to people that I hadn't tested properly myself. Considering majority of users copy-paste the examples and expect them to be working without issues - I believe this is better.
Redis : I've added it on all the places it was missing. It should be there, all it does is speeding up nextcloud in general.
Logging: Added a limit of 50 megabytes of log files on all containers. Recently I noticed that some instances had their log files grown to some absurd sizes (uptime for more than year without updates/downtime) and I've added that to the examples as well, for the same reason i changed the mariadb version. On few places(proxy containers, to be precise) I have intentionally added the logging driver to be
syslog
, because it works better with fail2ban on the host machine, it can't do any harm and it can hint to semi-sleeping sysadmins that the machine is under bots attack trying to log-in.EDIT:
nginx.conf
and fixed the cache control according to the official NC docs. I thought it was a typo, turns out it was quite cheeky way of adding it properly. Now it matches the official docs.closes #2200