502 Bad Gateway after Fresh Install #2939
Comments
|
I have the same problem, culprit seems to be that database login does not work (hence the |
|
Tested it on two new servers right now and cannot replicate it. You should try |
|
having the same issue but by running the update.sh script |
|
New installation No. 3 (edit: on Debian 10 btw.) ... Works. I also run update.sh multiple times on multiple machines and cannot replicate that on any supported mailcow installation. |
|
Hmm strange, I just tried again and still same problem: I can confirm hat $MYSQL_USER and $MYSQL_PASSWORD is the same as DBNAME and DBPASS in mailcow.conf MySQ container logs: Any ideas how to further debug this? |
|
Don't know. I cannot fix what does not fail on various installations of my machines or any machine in support. :/ Try to reset your SQL passwords. Or try to repair the table, perhaps it is broken. That would not explain the broken installation, but your upgrading problems. |
|
ran |
|
I don't think much changed. It is very important you wait a few minutes, if you run into the 502 page. We should change the default 502 to a "please wait" screen. :) |
well, I waited at least 24 hours :P - but a |
|
Running Mailcow in a LXC container was causing this problem for me. I now installed Mailcow in a KVM and it works perfectly fine now. |
|
I think the problem is related to MariaDB/mariadb-docker#262 |
|
Their logs read "mysqld 10.2.27" |
|
Yes, but I see in the log there is missed |
|
Ah, I understand. :) Not sure if that's the reason though: Affects Version/s: 10.4.3, 10.1.38, 10.2.22, 10.3.13, 10.0.38-galera I setup new mailcows from time to time and did not run into that problem. :( I think we can close this issue. We upgraded to 10.3 (whereas 10.3.18 is the latest) and should check if that errors persists. Feel free to reopen the issue then. |
|
Same thing happened to me today after running update.sh. |
|
Logs? |
|
I can confirm this problem still exists with a fresh installation (from today). I guess this part of the log describes the issue: I am running this on a fresh installation on debian 10 buster. |
|
I installed it a few minutes ago without a problem. |
|
The mailcow system is installed on a virtual machine with one complete IP address being redirected towards it. Also I have 2 virtual networks connected to it, one is isolated. Other than that I cannot think of any reason why I have a different setup than you. Are there any other logfiles or system information I can provide to you? Is there a way to wipe the container and rerun the setup? |
|
I think that's related to MYSQL_INITDB_SKIP_TZINFO (and that being empty, which is fine). I can try to move that part to another stage. Hm. Don't know if we reaaaally need it anyway. I will figure it out. |
|
Please try to change docker-compose.yml like this: MYSQL_INITDB_SKIP_TZINFO is new and set to 1 here. We don't use the TZINFO anywhere. Not sure about Roundcube etc. ... we could still import the zones at a later stage as mentioned previously. Please try and report back. :) |
|
No change :( |
|
Nothing else changed in the configs |
|
You need to run "docker-compose down -v" and "docker-compose up -d". The volume needs to be purged before testing. |
|
Can you try again and tell me if it still fails? If it does, please send me a mail to [email protected] |
|
That did it, thank you. I ran update.sh before thinking that would reset the container. |
|
I'm getting this even with docker-compose down -v and then docker-compose up -d. I've even added --rmi all to remove all images on the down and it doesn't work. It worked the first time, but then on restart it started failing even with defaults. The logs that it refers to just says "waiting for sql" endlessly for over 24 hours. I purged all of the volumes and manually deleted the contents of anything left. The only changes to my conf is disabling lets encrypt because I use pfsense to generate my lets encrypt certs and have a script that copies them in (and they're full 777) and I changed smtp to 2225 instead of 25 because of a conflict with something else using postfix on port 25. What other logs do you need? |
|
Having the same issue as @JohnGalt1717 :( |
can't reproducte. This is not a support ticket system. |
|
Hello, just want to remark that I just tried to install mailcow-dockerized (master branch, commit It started working after I removed the line from I removed that and now I get the login screen instead of error 502. |
|
We initialize the tzdata always via PHP-FPM. No matter what you set it to. You probably made a mistake somewhere along your tries or did not properly cleanup the volumes (I always read "freshly installed" but everyone seems to forget the volumes). It should not happen and you should really check your installation. Make sure you are not using ultra slow disks. |
|
I shutdown mailcow with I just tried out again with a "fresh" install (including checking that the volumes were not left behind!) and it worked (well, kind of, I could log-in), so it's not the TZ info. Thanks for chiming in! |
Prior to placing the issue, please check following: (fill out each checkbox with a
Xonce done)Description of the bug: What kind of issue have you exactly come across?
I got a new Server and did a fresh install of both OS and mailcow (tested with Debian 9.11 and Debian 10.1). I did it like ... 6 times now and everything is ending in an 502 Bad Gateway output of nginx so i couldnt reach the UI a single time. I read a few of the other 502 Issues in here but none of the hints was working for me.
Reproduction of said bug: How exactly do you reproduce the bug?
After that i try to open mail.xxxxxx.de and i get the 502 Bad Gateways Error
I have tried or I do... (fill out each checkbox with a
Xif applicable)System information
Further information (where applicable):
docker version)docker-compose version)Further notes:
git diff origin/master, any other changes to the code? If so, please post them.iptables -L -vn,ip6tables -L -vn,iptables -L -vn -t natandip6tables -L -vn -t natdocker exec -it $(docker ps -qf name=acme-mailcow) dig +short stackoverflow.com @172.22.1.254(set the IP accordingly, if you changed the internal mailcow network) anddocker exec -it $(docker ps -qf name=acme-mailcow) dig +short stackoverflow.com @1.1.1.1- output? Timeout?docker exec -it $(docker ps -qf name=acme-mailcow) dig +short stackoverflow.com @172.22.1.254
---> 151.101.1.69
151.101.65.69
151.101.129.69
151.101.193.69
docker exec -it $(docker ps -qf name=acme-mailcow) dig +short stackoverflow.com @1.1.1.1
---> 151.101.129.69
151.101.193.69
151.101.1.69
151.101.65.69
General logs:
NGinx:
This my netstat:
another thing i see is:
the mysql log shows:
The text was updated successfully, but these errors were encountered: