-
Notifications
You must be signed in to change notification settings - Fork 645
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
[3.0] Display worker CPU and memory utilization in supervisor list #589
Conversation
I feel like "CPU Threads" is going to be a bit confusing for people. I'm not used to seeing that as a metric. |
It's what is normally implied by UNIX systems when they show "CPU" (see Trying to normalize the number to get total CPU usage is non trivial on virtualized systems where not all resources may be available to the running container. |
This works on Windows, Mac, and Linux? |
Horizon 3.0 as is doesn't work on Windows, as it depends on This change simply parses the output from The UI of cause displays correctly on Windows as before. |
@driesvints I see you released |
@JayBizzle that's odd. I see @themsaid compiled it here: 8ee690a So it should be working afaik... |
Weird, can't get it to work after updating....let me try again... |
Nvm. Seems the pr was merged after it. I'll release a patch version with the compiled assets 👍 |
@JayBizzle released v3.2.1 |
Yep, just tested locally, had to rebuild the assets 👍 |
Updated to v3.2.1, all working as expected 👍 |
Great work @AJenbo, love this addition to Horizon 🔥 |
Thanks |
Very nice @AJenbo ! |
If the front end is running in a different machine from the queue processor one, the stats are not showing. Maybe saving the stats in redis should solve the issue. ;) |
I have tested it on systems where the frontend is running in a different machine. And the results are saved in Redish. @andreladocruz Could you describe more accurately exactly what you are experiencing? Did you restart your supervisor after updating the package? |
@AJenbo, we have our frontend running in google appengine, our queue processor running in a Computer Engine machine, and our redis is a Memorystore stack. Just updated horizon to 3.2.1 version and can´t any information about cpu and memory usage: That´s what I´m talking about. If the queue processor is in another server rather then the frontend one, no information is shown. |
Since the columns labels are entirely missing your browser is still displaying Horizon 3.2.0 or older, maybe you have some caching that hasn't been cleared after the deployment or your AppEngine wasn't updated properly. This has nothing to do with running on multiple servers. |
Please remember to follow the upgrade guide and re-publish the assets when updating horizon: https://laravel.com/docs/5.8/horizon#upgrading |
@AJenbo and @driesvints , thanks for reminding to publish the assets. =P Everything is working okay. Sorry about that. |
I experience a performance hit (+5-6% CPU usage) when upgrading from 3.1.2 to 3.2.1. Anybody else? I opened an issue: #601 |
Since this PR was causing quite some bugs for others and also introduced a performance hit we've decided to revert it entirely. We also won't be re-considering a new PR anymore. We'd like to encourage you to use other tools to track and measure CPU/Memory consumption and just keep Horizon to its core features. |
So here is my first PR in relation to #586
This adds information about how much system resources (CPU and memory) the workers are currently consuming:
![image](https://user-images.githubusercontent.com/204594/57295832-a460f380-70cb-11e9-855c-552f17457024.png)
I have tried to communicate this in the most user friendly way possible, but here are some additional details: