> Hi All,

> I'm trying to implement MMS on a sizeable platform with multiple,
> firewalled clusters and have run into a problem.

> The scenario is this: I've installed mms-agent on at least one host in
> each of the clusters (happens to be the MongoDB server but I know it
> doesn't matter which) and set them up in the MMS control panel.  However
> I'm getting constant alerts from MMS telling me my MongoDB servers are
> down, which they aren't.

> After some time debugging I've discovered the problem is that MMS's choice
> of what agent to monitor a server with seems to be pretty random - it
> sticks with an agent until it goes down then moves on to the next
> available.  The problem is that most of these agents are firewalled and do
> not have access to most of the MongoDB servers I would like to monitor.

> Does anybody know of a way to persuade MMS to use the appropriate agent to
> access each server?  I know we could open up firewall ports between all the
> clusters but this would rather defeat our security model!

> All the best,
> Alex.

> If you have multiple environments that cannot communicate with each other,
> the recommend solution is to create multiple groups. Each group would have
> a different agent (with different keys in settings.py - i.e., you have to
> download an agent for each the new groups).

> Only one agent is active at a time, the rest are hot-standbys -
> http://mms.10gen.com/help/single/#monitoring-architecture

> See this page for more information:
> http://mms.10gen.com/help/single/#multiple-environments

> Let us know if that solves your problem or if you have additional
> questions.

> On Thursday, September 27, 2012 5:48:38 AM UTC-4, alexmock wrote:

>> Hi All,

>> I'm trying to implement MMS on a sizeable platform with multiple,
>> firewalled clusters and have run into a problem.

>> The scenario is this: I've installed mms-agent on at least one host in
>> each of the clusters (happens to be the MongoDB server but I know it
>> doesn't matter which) and set them up in the MMS control panel.  However
>> I'm getting constant alerts from MMS telling me my MongoDB servers are
>> down, which they aren't.

>> After some time debugging I've discovered the problem is that MMS's
>> choice of what agent to monitor a server with seems to be pretty random -
>> it sticks with an agent until it goes down then moves on to the next
>> available.  The problem is that most of these agents are firewalled and do
>> not have access to most of the MongoDB servers I would like to monitor.

>> Does anybody know of a way to persuade MMS to use the appropriate agent
>> to access each server?  I know we could open up firewall ports between all
>> the clusters but this would rather defeat our security model!

>> All the best,
>> Alex.