-
Notifications
You must be signed in to change notification settings - Fork 185
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
Deployment getting stuck #333
Comments
Here is some log I found into the failed jobs table.
12h later, the deployment is still running. I thought there were a timeout after which the task is cancelled. |
Doing more testing, I reverted back to NodeJS 5.8.0.
The task he's getting stuck on is basically a If you have any idea, it would be great. |
More debug log, hope it helps:
|
"A queued job has been attempted too many times" hmm, what is your supervisor config setup? It should only try a job once |
Oh actually looking at the call to markJobAsFailedIfAlreadyExceedsMaxAttempts in the log it has only tried once. So, are you saying it has timed out but hasn't marked it as failed? You can clean it up with |
Thanks @REBELinBLUE, I have cleaned up many times and retried in different conditions. On the main project's page, I can see that the deployment is "Failed". |
hmm, OK I'll have to look, guess I have broken deployer/app/Console/Commands/ClearStalledDeployment.php Lines 59 to 84 in 70779a6
|
hmm, can you stop the queue listener/supervisor, run the cleanup command, run |
Which version of node & npm is on the server you are trying to deploy to? |
Hello @REBELinBLUE I did clean the queue many times, restarted supervisor and beanstakld but nothing did it. Still stuck I don't get it . Thanks |
I'm still trying to recreate this, you can fix the one which is marked as "running" by editing the something like (assuming you don't have any currently running deployments UPDATE server logs SET status = 3 WHERE status = 2; |
Are you sure your npm install command isn't prompting for input? What exactly is the command you have configured for that step |
Hello, The exact command is:
I don't know whether there is a prompt or not. The deployer doesn't display the output button. Nothing in the server_log table either (NULL). Running the same command manually works of course... |
Can you post your packages.json? |
Sure
|
I'm getting really confused now : I got rid of the NPM install step and it's getting stuck sooner now, on the composer install:
It's been 20min, still stuck on Twig. |
Confirmed into the failed_jobs table :
@REBELinBLUE what should I do? I don't understand why everything fails now. It used to work really fine before. |
argh, what the actual hell! Looking at the composer output, could you try loading a url via CURL on the server Something like I notice all the packages it installs are loaded from cache so maybe it is getting stuck when it needs to download something (which could also happen with npm as well) |
Thanks. CURL did download the JSON file, no problem here. |
Could you try editing |
Any luck? |
Hello @REBELinBLUE, my bad, I did not get your last message so I didn't try this. |
Sorry to hear that, I just really don't know what is happening without more output. Only other thing I could think is that it is related to #327 but there would be an error in the log about the job being too big |
I have the exact same issue.
And jobs failing with the following error message
This started happening since I added following commands before activating a new release
Other projects without npm are working fine. |
Could you try switching to master and see if that fixes it (running composer install --no-dev and npm install --production and php artisan app:update as if you were installing a new version)? |
The new version is working fine. Thanks @REBELinBLUE |
great to hear. Interesting that it was causing different output in the log for you (and presumably it was the same thing for @axeloz), everyone else it was giving a message saying |
Oops, well ... My mistake, I have been too impatient :) |
Ha ha, no problem of course, when you need something to work you need it to work. Surprised Laravel doesn't catch the error to be honest, may have to look at seeing if I can catch it myself as there is no reason it should cause a fatal error. Silly thing it, it only came up because I fixed a bug. Previously if you were viewing the server log and it changed you didn't see the change without closing the window even though you were supposed to, I think the broadcasting so that correctly sent the change but in the process the log often ended up being bigger than the default limit for a beanstalkd job and it appears my dev server has it set higher than the default so it didn't happen for me. |
Before submitting your issue, please make sure that you've checked all of the checkboxes below.
php -v
composer install --no-dev
from the root of your installation.npm install --production
from the root of your installation.rm -rf bootstrap/cache/*
from the root of your installation.Expected behaviour
The deployer should run properly
Actual behaviour
The deployer seems to run VERY slowly, so slowly that I don't know if it still runs or is stuck.
I have been waiting very long (1h or more). It is getting stuck on the "npm install" task.
This project was working yesterday and the days before. I have upgraded Node in the meantime. I have deleted the node_modules folder and run
npm install --production
again.Steps to reproduce
No idea
Environment info
Operating System: Ubuntu 14
PHP Version: 7.1
Node Version: 7.7.1
Database System: MySQL
Database Version: 5.5.50-38.0
Logs (see storage/logs/) or other output that would be helpful
(If logs are large, please upload as attachment).
The text was updated successfully, but these errors were encountered: