Baikal Version 0.9.1

[tanrak]

Just for your information. There is a new version of Baikal. Could you also update your image? Many thanks!

[ckulka]

Thanks, unfortunately my application to the Docker Open Source Program is still open (openend it about half a year ago), so I can't build and publish a new version right away. I'm currently working on migrating the build process over to Github Actions, I hope I have a solution for the multi-arch build soon.

[ckulka]

So, good news: I figured out multi-arch builds with GitHub Actions, should be there tomorrow.

[ckulka]

I just pushed 0.9.1 as ckulka/baikal:experimental and ckulka/baikal:experimental-nginx.

I don't run Baikal myself anymore, can you please give them a try and see if they work?

If all goes well, I'll release them.

[tanrak]

@ckulka Thanks for the update. After a few test everything seems to work without any problems.

I don't run Baikal myself anymore, can you please give them a try and see if they work?

May I ask what software you are using now?

[ckulka]

Looks like everything published fine now and I can close the issue: 0.9.1 is available, thanks for testing it 🎉

Regarding me and Baikal, you'll be disappointed: I cycled through a lot of devices, laptops and servers a few years ago, running my own addressbook and calendar just consumed too much time. At the same time the announcement from the (then) maintainer came out that Baikal won't be maintained anymore, so I switched to Google (given Android).

Now with the new maintainers and Docker being a lot more mature/easier to orchestrate I might come back again, but other hobbies/projects keep me busy enough for now.

[tanrak]

Thanks for the information and yes, there are always things that are more important than others.

[mstilkerich]

Hi,

thanks for providing this docker image which I use in my CI environment to test interoperability with baikal.

With version 0.9 of the docker images, you switched to PHP 8.1. PHP 8.1 deprecates passing null values to internal functions with non nullable parameters. sabre/vobject contained as part of Baikal 0.9 has issues with that. So I would suggest to publish the images for 0.9 with php 8.0.x, not 8.1.

Baikal 0.8.0 contains the same issues, but since the docker image contains php 8.0 it works fine.

You can find evidence here:
https://github.com/mstilkerich/carddavclient/runs/4812843591?check_suite_focus=true

And from the Baikal log:

[php:notice] [pid 21] [client 192.168.10.1:44094] ErrorException: strtoupper(): Passing null to parameter #1 ($string) of type string is deprecated in /var/www/baikal/vendor/sabre/vobject/lib/Component.php:241\nStack trace:\n#0 [internal function]: Baikal\\Framework::exception_error_handler(8192, 'strtoupper(): P...', '/var/www/baikal...', 241)\n#1 /var/www/baikal/vendor/sabre/vobject/lib/Component.php(241): strtoupper(NULL)\n#2 [internal function]: Sabre\\VObject\\Component->Sabre\\VObject\\{closure}(Object(Sabre\\VObject\\Property\\FlatText))\n#3 /var/www/baikal/vendor/sabre/vobject/lib/Component.php(242): array_filter(Array, Object(Closure))\n#4 /var/www/baikal/vendor/sabre/vobject/lib/Component.php(464): Sabre\\VObject\\Component->select('EMAIL')\n#5 /var/www/baikal/vendor/sabre/dav/lib/CardDAV/Plugin.php(489): Sabre\\VObject\\Component->__isset('item1.EMAIL')\n#6 /var/www/baikal/vendor/sabre/dav/lib/CardDAV/Plugin.php(433): Sabre\\CardDAV\\Plugin->validateFilters('BEGIN:VCARD\\r\\nVE...', Array, 'anyof')\n#7 /var/www/baikal/vendor/sabre/dav/lib/CardDAV/Plugin.php(194): Sabre\\CardDAV\\Plugin->addressbookQueryReport(Object(Sabre\\CardDAV\\Xml\\Request\\AddressBookQueryReport))\n#8 /var/www/baikal/vendor/sabre/event/lib/WildcardEmitterTrait.php(89): Sabre\\CardDAV\\Plugin->report('{urn:ietf:param...', Object(Sabre\\CardDAV\\Xml\\Request\\AddressBookQueryReport), 'addressbooks/ci...')\n#9 /var/www/baikal/vendor/sabre/dav/lib/DAV/CorePlugin.php(685): Sabre\\DAV\\Server->emit('report', Array)\n#10 /var/www/baikal/vendor/sabre/event/lib/WildcardEmitterTrait.php(89): Sabre\\DAV\\CorePlugin->httpReport(Object(Sabre\\HTTP\\Request), Object(Sabre\\HTTP\\Response))\n#11 /var/www/baikal/vendor/sabre/dav/lib/DAV/Server.php(472): Sabre\\DAV\\Server->emit('method:REPORT', Array)\n#12 /var/www/baikal/vendor/sabre/dav/lib/DAV/Server.php(253): Sabre\\DAV\\Server->invokeMethod(Object(Sabre\\HTTP\\Request), Object(Sabre\\HTTP\\Response))\n#13 /var/www/baikal/vendor/sabre/dav/lib/DAV/Server.php(321): Sabre\\DAV\\Server->start()\n#14 /var/www/baikal/Core/Frameworks/Baikal/Core/Server.php(119): Sabre\\DAV\\Server->exec()\n#15 /var/www/baikal/html/dav.php(69): Baikal\\Core\\Server->start()\n#16 {main}

I will prepare a pull request for sabre/vobject, but anyway the fix would become only part of a future baikal release.

[ckulka]

Thanks @mstilkerich - btw, nice to see someone using a CI pipeline with Baikal, I always wanted to do something similar for the Baikal Docker image!

I'll open an incident over at sabre-io/Baikal later today (unless you beat me to it of) because their release 0.9.0 (https://github.com/sabre-io/Baikal/releases) says it now supports PHP 8.1 😕

Do you know if this would break Baikal for regular users, or is this "just" for something special in carddavclient? Asking because I would then re-release with PHP 8.0 as soon as I can.

[mstilkerich]

Hello,

it happens when an addressbook search is done that includes a group filter. It is probably not a very common action, but nevertheless with PHP 8.0 everything works just fine.

[mstilkerich]

Created a PR here: sabre-io/vobject#561

[Lithimlin]

Since the new version I'm regularly getting gateway timeouts when querying my calendars or any admin page:

baikal  | 172.30.0.3 - - [07/Apr/1982:12:45:52 +0000] "GET /res/core/BaikalAdmin/main.js HTTP/1.1" 304 0 "https://<mydomain>/admin/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 1982/04/07 12:45:52 [error] 21#21: *128 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 172.30.0.3, server: _, request: "GET /admin/?/users/ HTTP/1.1", upstream: "fastcgi://unix:/var/run/php-fpm.sock", host: "<mydomain>", referrer: "https://<mydomain>/admin/"
baikal  | 172.30.0.3 - - [07/Apr/1982:12:45:52 +0000] "GET /admin/?/users/ HTTP/1.1" 504 167 "https://<mydomain>/admin/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"

Does anyone know something about this?

[mstilkerich]

Since you are using php fpm, any errors in the fpm logs?

[ckulka]

I published images with Baikal 0.9.1 and PHP 8.0, @mstilkerich and @Lithimlin can you please see if your issues were resolved by reverting to the older PHP version?

• ckulka/baikal:experimental-apache-php8.0
• ckulka/baikal:experimental-nginx-php8.0

[mstilkerich]

Hi, thanks for providing this update! My test suite passes all tests! (I only tried the apache image, but I wouldn't think it makes a difference)

[Lithimlin]

Now I always get a timeout. Here are my logs:

baikal  | 1960/08/02 06:04:48 [error] 21#21: *1 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 172.30.0.2, server: _, request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/var/run/php-fpm.sock", host: "<my-domain>"
baikal  | 172.30.0.2 - - [02/Aug/1960:06:04:48 +0000] "GET / HTTP/1.1" 504 167 "-" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 172.30.0.2 - - [02/Aug/1960:06:04:48 +0000] "GET / HTTP/1.1" 504 167 "-" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 1960/08/02 06:04:48 [error] 21#21: *3 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 172.30.0.2, server: _, request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/var/run/php-fpm.sock", host: "<my-domain>"
baikal  | 1960/08/02 06:04:48 [error] 21#21: *5 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 172.30.0.2, server: _, request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/var/run/php-fpm.sock", host: "<my-domain>"
baikal  | 172.30.0.2 - - [02/Aug/1960:06:04:48 +0000] "GET / HTTP/1.1" 504 167 "-" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 1970/03/03 11:40:09 [error] 21#21: *7 open() "/var/www/baikal/html/favicon.ico" failed (2: No such file or directory), client: 172.30.0.2, server: _, request: "GET /favicon.ico HTTP/1.1", host: "<my-domain>", referrer: "https://<my-domain>/"
baikal  | 172.30.0.2 - - [03/Mar/1970:11:40:09 +0000] "GET /favicon.ico HTTP/1.1" 404 153 "https://<my-domain>" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 1960/08/02 06:04:48 [error] 21#21: *8 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 172.30.0.2, server: _, request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/var/run/php-fpm.sock", host: "<my-domain>"
baikal  | 172.30.0.2 - - [02/Aug/1960:06:04:48 +0000] "GET / HTTP/1.1" 504 167 "-" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 1970/03/03 11:40:09 [error] 21#21: *10 open() "/var/www/baikal/html/favicon.ico" failed (2: No such file or directory), client: 172.30.0.2, server: _, request: "GET /favicon.ico HTTP/1.1", host: "<my-domain>", referrer: "https://<my-domain>/"
baikal  | 172.30.0.2 - - [03/Mar/1970:11:40:09 +0000] "GET /favicon.ico HTTP/1.1" 404 153 "https://<my-domain>/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 1960/08/02 06:04:48 [error] 21#21: *11 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 172.30.0.2, server: _, request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/var/run/php-fpm.sock", host: "<my-domain>"
baikal  | 172.30.0.2 - - [02/Aug/1960:06:04:48 +0000] "GET / HTTP/1.1" 504 167 "-" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 1970/03/03 11:40:09 [error] 21#21: *13 open() "/var/www/baikal/html/favicon.ico" failed (2: No such file or directory), client: 172.30.0.2, server: _, request: "GET /favicon.ico HTTP/1.1", host: "<my-domain>", referrer: "https://<my-domain>/"
baikal  | 172.30.0.2 - - [03/Mar/1970:11:40:09 +0000] "GET /favicon.ico HTTP/1.1" 404 153 "https://<my-domain>/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 1960/08/02 06:04:48 [error] 21#21: *14 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 172.30.0.2, server: _, request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/var/run/php-fpm.sock", host: "<my-domain>"
baikal  | 172.30.0.2 - - [02/Aug/1960:06:04:48 +0000] "GET / HTTP/1.1" 504 167 "-" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 1960/08/02 06:04:48 [error] 21#21: *16 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 172.30.0.2, server: _, request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/var/run/php-fpm.sock", host: "<my-domain>"
baikal  | 172.30.0.2 - - [02/Aug/1960:06:04:48 +0000] "GET / HTTP/1.1" 504 167 "-" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"

cat /var/log/php8.1-fpm.log 
[02-Apr-1935 21:32:24] NOTICE: fpm is running, pid 14
[28-Mar-1970 06:37:52] NOTICE: ready to handle connections
[21-Nov-2003 04:13:52] NOTICE: systemd monitor interval set to 10000ms

And my docker-compose:

  baikal:
    #image: ckulka/baikal:nginx
    image: ckulka/baikal:experimental-nginx-php8.0
    container_name: baikal
    restart: always
    volumes:
     - ./volumes/baikal/config:/var/www/baikal/config
     - ./volumes/baikal/data:/var/www/baikal/Specific
    networks:
     - caddy-net

[mstilkerich]

@ckulka I now tried the nginx image as well - it still contains php 8.1?! (and the corresponding tests fail as expected)

Yep, this is because the workflow to build the 8.0 nginx image references the dockerfile for 8.1, not the 8.0 one.

Btw I could not reproduce the timeouts, the nginx image works as expected. I noticed in your logs that the system time seems wrong and weirdly inconsistent between the fpm log and the nginx log, also both run inside the same container.

[ckulka]

Indeed, my mistake: I built them locally for tests and didn't notice the Nginx PHP 8.0 GitHub Action used the wrong Dockerfile. Fixed that, ckulka/baikal:experimental-nginx-php8.0 now runs PHP 8.0.

[Lithimlin]

I still get the occasional timeout even with the new image.

docker logs:

baikal  | 172.30.0.2 - - [31/Oct/1914:16:14:56 +0000] "GET /admin/ HTTP/1.1" 200 2334 "-" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 172.30.0.2 - - [31/Oct/1914:16:14:56 +0000] "GET /res/core/BaikalAdmin/GlyphiconsPro/glyphpro.css HTTP/1.1" 304 0 "https://<my-domain>/admin/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 172.30.0.2 - - [31/Oct/1914:16:14:56 +0000] "GET /res/core/BaikalAdmin/GlyphiconsPro/glyphpro-2x.css HTTP/1.1" 304 0 "https://<my-domain>/admin/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 172.30.0.2 - - [31/Oct/1914:16:14:56 +0000] "GET /res/core/TwitterBootstrap/js/jquery-3.1.0.min.js HTTP/1.1" 304 0 "https://<my-domain>/admin/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 172.30.0.2 - - [31/Oct/1914:16:14:56 +0000] "GET /res/core/TwitterBootstrap/css/bootstrap-responsive.css HTTP/1.1" 304 0 "https://<my-domain>/admin/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 172.30.0.2 - - [31/Oct/1914:16:14:56 +0000] "GET /res/core/BaikalAdmin/Templates/Page/style.css HTTP/1.1" 304 0 "https://<my-domain>/admin/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 172.30.0.2 - - [31/Oct/1914:16:14:56 +0000] "GET /res/core/TwitterBootstrap/js/bootstrap.min.js HTTP/1.1" 304 0 "https://<my-domain>/admin/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 172.30.0.2 - - [31/Oct/1914:16:14:56 +0000] "GET /res/core/TwitterBootstrap/js/jquery.color-2.2.0.min.js HTTP/1.1" 304 0 "https://<my-domain>/admin/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 172.30.0.2 - - [31/Oct/1914:16:14:56 +0000] "GET /res/core/BaikalAdmin/Templates/Page/baikal-text-20.png HTTP/1.1" 304 0 "https://<my-domain>/admin/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 172.30.0.2 - - [31/Oct/1914:16:14:56 +0000] "GET /res/core/BaikalAdmin/main.js HTTP/1.1" 304 0 "https://<my-domain>/admin/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 172.30.0.2 - - [31/Oct/1914:16:14:56 +0000] "GET /res/core/BaikalAdmin/GlyphiconsPro/glyph2x-dark.png HTTP/1.1" 304 0 "https://<my-domain>/res/core/BaikalAdmin/GlyphiconsPro/glyphpro-2x.css" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"
baikal  | 1970/12/17 15:21:20 [error] 19#19: *12 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 172.30.0.2, server: _, request: "POST /admin/ HTTP/1.1", upstream: "fastcgi://unix:/var/run/php-fpm.sock", host: "<my-domain>", referrer: "https://<my-domain>/admin/"
baikal  | 172.30.0.2 - - [17/Dec/1970:15:21:20 +0000] "POST /admin/ HTTP/1.1" 504 167 "https://<my-domain>/admin/" "Mozilla/5.0 (X11; Linux x86_64; rv:96.0) Gecko/20100101 Firefox/96.0" "77.181.218.126"

The php log is still as helpful as before:

/var/log# cat php8.0-fpm.log 
[16-Mar-1935 06:17:04] NOTICE: fpm is running, pid 14
[26-Mar-1970 21:36:28] NOTICE: ready to handle connections
[03-Aug-1948 22:30:24] NOTICE: systemd monitor interval set to 10000ms

[mstilkerich]

Hello, my test suite passes with the new image, I don’t see these timeout issues. Why are the timestamps in your logfiles jumping by decades for entries that should be logged directly in a row?!

[Lithimlin]

I'm not sure. Something seems to be going significantly wrong but I can't figure out what it is. Could it be a problem with a timezone variable?

[mstilkerich]

@ckulka From the Baikal issue linked above it appears these issues are also occurring in Baikal’s own code. And yes, apparently also affecting regular users.

[ckulka]

Figured out how I will create new releases that replace existing Baikal-versions, all is prepared to release the PHP 8.0 variants.

Thanks @mstilkerich, good to see the tests pass, I plan to publish the new images tomorrow night/day after tomorrow unless I hear of issues (except for @Lithimlin).

Hi @Lithimlin, I'm still trying to think of a reason why your time inside the container is all over the place, but I can only think of something caused by the host. Can you please run the following commands a couple of times and show the output?

docker run --rm -it ckulka/baikal:0.9.1 date
docker run --rm -it ckulka/baikal:0.9.1-nginx date

docker run --rm -it ckulka/baikal:0.8.0 date
docker run --rm -it ckulka/baikal:0.8.0-nginx date

[Lithimlin]

The date seems to be messed up with the 0.9.1 release no matter the flavor of image:

~ $ docker run --rm -it ckulka/baikal:0.9.1 date
Thu Jan  1 00:00:00 UTC 1970
~ $ docker run --rm -it ckulka/baikal:0.9.1-nginx date
Thu Jan  1 00:00:00 UTC 1970
~ $ docker run --rm -it ckulka/baikal:0.8.0 date
Wed Jan 26 10:10:14 UTC 2022
~ $ docker run --rm -it ckulka/baikal:0.8.0-nginx date
Wed Jan 26 10:10:18 UTC 2022

[ckulka]

I found someone with a very similar sounding issue at Date is wrong in container (linuxserver.io) who points to this solution: https://docs.linuxserver.io/faq#my-host-is-incompatible-with-images-based-on-ubuntu-focal.

Can you give that a try?

[ckulka]

Just FYI, I just re-published 0.9.1 including the new PHP 8.0 images as 0.9.1+php8.0.

The new PHP 8.0-based images are:

• 0.9.1-php8.0, 0.9.1-apache-php8.0, apache-php8.0, latest-php8.0
• 0.9.1-nginx-php8.0, nginx-php8.0

The existing PHP 8.1-based images have been updated with the latest base images:

• 0.9.1, 0.9.1-apache, apache, latest
• 0.9.1-nginx, nginx

[Lithimlin]

I found someone with a very similar sounding issue at Date is wrong in container (linuxserver.io) who points to this solution: https://docs.linuxserver.io/faq#my-host-is-incompatible-with-images-based-on-ubuntu-focal.

Can you give that a try?

Ah, it seems that my current Raspbian is not supported anymore then.
I'm currently in the process of setting everything up so I can raze the Pi and install everything again. I'll let you know if an update helps.

[Lithimlin]

I just double-checked and interestingly enough I'm already on Bullseye.
Just to make sure I updated everything but I still get the wrong date.
So I will wait until I updated the Pi and everything and let you know again whether things are working.

[ckulka]

@Lithimlin I moved your issue to #72 since it (probably) has a different cause and to avoid flooding OP with updates unrelated to the original request.

[mstilkerich]

Thank you, the new image works for me, so the issue is resolved from my side.