Thank you, Marcus. I'll give it a shot!

tlj
-----Original Message-----
From: [email protected] 
[mailto:[email protected]] On Behalf Of Marcus Gatt
Sent: Wednesday, October 03, 2012 10:04 PM
To: [email protected]
Subject: [BULK] Re: [Supervisor-users] [supervisor-users] managing/monitoring 
distributed

We built a very basic web interface for multiple Supervisor servers using the 
XML-RPC interface. You can get it here:

https://github.com/luxbet/supervisorui

Functionality is mostly limited to start/stop. It is PHP based though (sorry, 
we are a PHP shop for the most part).

--
Marcus


On Oct 3 02:13:23 UTC 2012, Timothy Jones <Timothy.Jones at syniverse.com>  
wrote:

Today, the same suggestion came up where I work, about navigating multiple 
supervisor instances through one web interface.

I am familiar with the XMLRPC API and that works well for shell scripts, but we 
are interested in a way to do this from a webserver as well.

A very basic form of inter-instance navigation might be to include a HTML 
snippet that gets included in each process-listpage that is displayed.  This 
snippet would include HTML <A HREF> tags to the other supervisor instances.

A better form might be for each supervisor instance to contact all the others 
through the XMLRPC interface and show all the processes on one page.  The 
hyperlinks for the start/stop/tail functions would invoke those actions on the 
other supervisor instances directly. The hostname in the browser's URL bar 
would change in this case.

What do you think?



tlj

On Oct 2, 2012, at 8:50 PM, "Phil Whelan" <phil123 at
gmail.com<mailto:phil123 at gmail.com>> wrote:

Hi Spencer,

Have you looked at the XMLRPC API for Supervisord?
http://supervisord.org/api.html

With this it's quite easy to remotely query status and stop|start processes. 
You can do this for multiple instances of supervisord quite easily.

Cheers,
Phil


On Tue, Oct 2, 2012 at 3:08 PM, Spencer Liechty <spencer.liechty at 
adbucket.com<mailto:spencer.liechty at adbucket.com>> wrote:
Hey all,

I've been using supervisord to run persistent scripts on my web application 
servers, and am wondering what best practices exist for a couple of challenges 
I'm facing.

1. Process Logging: I'm currently logging all process logs locally to 
/var/log/supervisor/ on my app servers. Does supervisor suggest/provide a 
standard way to retrieve those logs from my various servers? I suppose I could 
scp that directory daily/weekly/etc, but am wondering if there exists a 
canonical supervisor-approved way of aggregating those logs.

2. Process Monitoring/Management: I know I can set up inet_http_servers on 
every web server I'm running and manage processes by simply navigating to each 
server's supervisorctl web interface, but I was wondering if there was a way to 
manage all my different supervisord servers from a central location. For 
example, I have 10 app servers running supervisord on each one, and would like 
to restart all processes on all servers. Any ideas on an easy/scriptable way to 
do this? Would y'all suggest thinking of dev op tools for activities such as 
this?

Thanks for the help!
Spencer
_______________________________________________
Supervisor-users mailing list
[email protected]
http://lists.supervisord.org/mailman/listinfo/supervisor-users
_______________________________________________
Supervisor-users mailing list
[email protected]
http://lists.supervisord.org/mailman/listinfo/supervisor-users

Reply via email to