![]() |
![]() |
Subject: | Start, stop, and status applications |
Author: | [Not Specified] |
Posted: | 2014-03-17 08:38 |
We are considering Zenoss and I'd like feedback on suitability to start, stop and status applications. Here's the functionality I'm looking for:
1. It would need to handle dependencies for starting applications (app A requires app B which requires app C)
2. I'm hoping an agents can handle actually starting an application for a specific user ID
3. The display should include a color status (green - started, red - stopped, yellow - crashed) for a logical collection of servers, a single server, and each application. For each application, details about the status (PID, start time, etc.) would be displayed along with the option to start or stop it (if they are in a specified LDAP/Active Directory group).
I haven't seen any screen shots that would work for the display and figure it would need to be heavily customized. Is this possible, and well documented
I welcome any input
Subject: | Possible but would require customization to do what you want |
Author: | [Not Specified] |
Posted: | 2014-03-17 09:19 |
Zenoss' out of the box process monitoring consists of defining rules for what processes you want to monitor (based on regex statements). Once those rules are in place it auto-discovers processes running on all devices and monitors them. It does not however keep track of pids. It does keep track of the status of processes as well as how many of each are running, the amount of CPU, and the amount of memory consumed by each of them. It also knows whether they are running or not and generates an alert if the process fails. It also has Windows Services monitoring support available that does much the same thing but for windows services. There is a trigger system in Zenoss where you can tell it to run commands when it sees a certain event (and can pass properties from the event to the command). You can make use of things like winexe or ps tools on linux to restart windows processes/services, and you can make use of simple ssh scripts to do it on unix hosts.
As far as process requirements/dependencies, that could partially be accomplished with event transforms (which allow you to develop complex event processing rules) and partially via custom scripts.
Subject: | 1. Impact would be perfect |
Author: | Jay Stanley |
Posted: | 2014-08-13 19:02 |
1. Impact would be perfect for this.
2. Sure, Zenoss can use check_nrpe and NRPE and nsclient can be setup to run commands.
3. It has colors - Red, Yellow, Orange
< |
Previous How to use heartbeat feature for own events? |
Next zenoss 4.2.4 Network map |
> |