-
-
Notifications
You must be signed in to change notification settings - Fork 119
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
No log line matching the '' filter #104
Comments
@ShlomiD83 can you please dump some logs below so I can take a look? |
@alexjustesen the issue is that the logs for the container are blank. I just tried to spin up a container after seeing the video from DBTech and I get the same "error" in the logs. I am running this in portainer on an Ubuntu 20.04 VM with 8 cores and 16GB RAM. The only thing that I have changed from the compose example is the volume and the port (as well). I have been looking forward to an updated version of what I have enjoyed. Keep up the great work and I will keep testing for you. |
I am following the example here: https://docs.speedtest-tracker.dev/getting-started/installation - Docker (MariaDB/MySQL) I did see the example on github and not sure why they are different but that is the one that was covered on DBTech's video. |
can you drop a link to that example in case I missed an old doc? |
which logs? It seems to be empty. |
I am receiving the same error above "No log line matching the '' filter". SQLSTATE[HY000] [2002] php_network_getaddresses: getaddrinfo for db failed: In Connector.php line 70: SQLSTATE[HY000] [2002] php_network_getaddresses: getaddrinfo for db failed: In Connector.php line 70: PDO::__construct(): php_network_getaddresses: getaddrinfo for db failed: Na s6-rc: warning: unable to start service startup: command exited 1 |
I've waited a few minutes and the logs started filling up too. |
the container takes forever to load up after restart/recreate. |
Are you sure you're using mine? Most start times are between 2 and 6s unless it's the first start when it creates a database for you. |
'ghcr.io/alexjustesen/speedtest-tracker:latest' That's the image I'm using, it's seems to be yours.. |
never saw this happening here. pretty quick actually. |
I've timed it at 8 minutes to load up. |
Houston, you have a problem... |
That's the only container that's acting up. |
Can you give me some details on your platform you're deploying on along with if you're using sqlite or another db? |
I'm running 66 containers...never took 8 minutes for one of them, and I have pretty "heavy" containers too. Take also into account you're the only one reporting this problem. What platform are you on? Can you try an older version tag too, just to verify. |
Raspberry pi 4 8GB, using sqlite. |
tried with "ghcr.io/alexjustesen/speedtest-tracker:v.0.1.0-beta". |
are you on sdcard? if you have mysql/mariadb, can you create the db and try with that configuration? |
no, I'm using USB flash drive for the OS/Portainer and all the container data is written to an external SSD connected via USB. |
are you sure data is being written to the external drive? can you post the docker-compose? Also, if you have the log of the container, just to check what's going on... |
yes, I can see database.sqlite written to the SSD. version: '3.3' |
_speedtest-tracker_logs (1).txt |
@ShlomiD83 I just released v0.2.0 which has arm v7 support (better rpi performance) let me know if this latest build changes anything for you. |
@alexjustesen when I see the "Starting the queue worker..." can I assume the startup phase is completed and services are up? Or is there stuff still going on in the background? It would be good to be sure that last line of the log means startup is completed. Also: the log doesn't provide timestamps...Portainer does luckily, otherwise we couldn't measure anything... |
This is your problem, but there are not enough details to debug it:
I also see this:
You might want to map the certs folder for SSL, otherwise it will autogenerate it at every startup of the container, but on the RPi I would avoid it.
|
which certs to map? I didn't create any SSL certs. |
same behavior with v0.2.0 unfortunately. |
The container creates it for SSL connection.
|
so that's the high CPU reason? or a problem with the web server? |
No, I explained why in previous post. :) It's a tiny optimization, it's not the problem you're having. The problem you're having is the error in the log regarding php-fpm described in previous post. |
OK, how can we debug it further? |
@alexdelprete @alexjustesen with v0.2.0 I don't see the PHP error in the logs but the container still takes time to load up. |
we need a debug level in the log...:) let's wait for @alexjustesen to guide us on how to better debug this. If you google...you'll find complaints about newer php-fpm versions causing performance issues on RPi. But we're lucky because in your case there's an ERROR in the log...but the detail has been truncated. |
I was googling, and read that there might be some caching going on to the boot drive (php-fpm), it means your usb drive. Why didn't you use the wdred for boot + data? RPi supports booting from USB. |
the wdred was bought after the system was running for a few weeks, didn't want to reinstall everything. |
You can clone the USB to wdred. I've used this in the past, it works fine: https://github.com/billw2/rpi-clone The problem is for example that many apps use the system drive for various things, caching, /tmp, etc. so you can always face some issues. |
OMV web UI didn't work after cloning, I've tried with other drives before. |
that's the first time this kind of thing happens, I've spun up hundreds of docker containers on this RPi with no issue, so I doubt that the USB drive is the culprit here. |
I've used it 2 times on two different RPi, an old RPi3 and a newer RPi4B, it always worked perfectly. Anyway, you should really think about getting rid of the USB drive for system. Choose your preferred method to migrate. |
on what basis you say that? if the container uses cache on the system drive, that would be your usb, and this is just one example. which containers in your setup are using php-fpm? google php-fpm and rpi, you'll find a lot of issues. good luck, I'll leave the rest to Alex. :) |
according to "/etc/php/8.1/fpm/php.ini" there is no caching happening. maybe that's not the issue? edit: restarting the container after it loaded up successfully for the first time takes seconds. |
The issue is your hardware setup, not the container, you're the only person having to wait 8mins to startup a container. Setup your RPi properly. Running docker on a usb key in 2022 is not recommended by anybody, particularly when you have an external fast device attached used only for data and not for booting OS. Do your migration and your problem will be solved, like it was for this guy: https://www.reddit.com/r/NextCloud/comments/hrdpib/nextcloud_on_raspberry_pi_4_with_mariadb/ But as I said, I'm out of this conversation, you don't want to listen, you just want to pretend the container has a bug, which has not. Goodbye, and good luck. |
@alexjustesen thanks for trying to assist, I'll just keep using "henrywhitaker3/speedtest-tracker" for now. |
I took your advise and moved my OS to an SSD drive, now the container loads up in just under a minute. |
Incredible how slow a USB drive is vs an SSD, isn't it? And now you also learned why booting the OS on the USB vs SSD is totally different, like I told you when you didn't want to listen. ;)
I use a Proxmox LXC container with docker on a Lenovo ThinkCentre M910x (Quad Core Intel Core i7-6700T), 32GB or RAM and a 1TB SSD. Can't really compare that to an RPi. And it's a desktop PC bought for $180. It runs a lot of stuff. Speedtest-Tracker starts in a few seconds, like other apps. |
Hi,
I've deployed the container on RPi 4 using docker compose with sqlite database.
The log shows only "No log line matching the '' filter".
The only thing I've changed is the port on the host side and the volume mapping.
Please advise.
The text was updated successfully, but these errors were encountered: