Hello all,
I’m migrating from vanilla Nagios to Check_MK (subscription OMD) I’m struggling to find the most efficient and scalable method of assigning custom Active HTTP checks to hosts. I would be grateful for any insight you might be able to share. Here are my constraints and objectives:
I have many dozens of web applications to monitor across various single instance nodes and load balanced farms (in which case I need to test the nodes and the VIP). Many hosts and VIPs are not dedicated to a single web app and instead host multiple virtual servers.
I need to check for a web application’s return string, but I can’t enforce a standard page to call or string to expect. It could be “index.cfm” with “alive” on one farm and “menu.maf” and “Edwards” on another farm, etc. At this point I’ve decided to create a separate active HTTP check for each web app, i.e. “HTTP-BLACKSMITH”, “HTTP-E1”, “HTTP-VLTRADER”, etc. This is how the services were named and defined in Nagios, and the assignment of services to hosts was via hostgroup.
In OMD I’m experimenting with a single on/off tag group for each service. After a creating my first dozen of these I’m seeing that it may not scale well to dozens in the GUI. I’d prefer to have a single tag group with a tag for each possible web app, but that would prevent me from assigning multiple tags to a single host. I’m beginning to wonder if auxiliary tags would help, but I think that might get a bit messy too.
Does anyone have a similar environment and a solution they are happy with? Thank you very much in advance.
Tony
Try using custom variable to pass all parameters specific to check.
extra_host_conf["_TESTURL"] = [
( “/login”, [“se-app01”] ),
]
extra_nagios_conf += r"""
define command {
command_name check-http command_line $USER5$/check_http
-t 20 -H $HOSTNAME$ -I $HOSTADDRESS$ -u “$_HOSTTESTURL$”
}
“”"
Thanks
Ashish
On Behalf Of Tony Adams
···
Hello all,
I’m migrating from vanilla Nagios to Check_MK (subscription OMD) I’m struggling to find the most efficient and scalable method of assigning custom Active HTTP checks to hosts. I would be grateful for any insight you might be able to share.
Here are my constraints and objectives:
I have many dozens of web applications to monitor across various single instance nodes and load balanced farms (in which case I need to test the nodes and the VIP). Many hosts and VIPs are not dedicated to a single web app and instead
host multiple virtual servers.
I need to check for a web application’s return string, but I can’t enforce a standard page to call or string to expect. It could be “index.cfm” with “alive” on one farm and “menu.maf” and “Edwards” on another farm, etc. At this point
I’ve decided to create a separate active HTTP check for each web app, i.e. “HTTP-BLACKSMITH”, “HTTP-E1”, “HTTP-VLTRADER”, etc. This is how the services were named and defined in Nagios, and the assignment of services to hosts was via hostgroup.
In OMD I’m experimenting with a single on/off tag group for each service. After a creating my first dozen of these I’m seeing that it may not scale well to dozens in the GUI. I’d prefer to have a single tag group with a tag for each
possible web app, but that would prevent me from assigning multiple tags to a single host. I’m beginning to wonder if auxiliary tags would help, but I think that might get a bit messy too.
Does anyone have a similar environment and a solution they are happy with? Thank you very much in advance.
Tony