cancel
Showing results for 
Search instead for 
Did you mean: 

Disable VIP whenever WAN/Internet failure occurs

Highlighted
Occasional Contributor

Disable VIP whenever WAN/Internet failure occurs

Hi all,

 

i have a setup of smtp servers being LB, and exchange will use the smtp vip for outgoing/sending emails. Everything works fine, but one of the possible failure that we can see is when there is an internet failures due to;

 

1) exchange does not see internet failures, the exchange servers would just sent to the smtp appliances

2) exchange can redirect to another smtp gateway, but only if smtp appliances is failed/vs disabled

3) smtp health monitors only smtp devices, not wan internet

 

so does anyone knows how to do this? i mean, how can we stop a VS, whenever internet failed a certain number of ping failed. i have 2 data centers with exact setup (diff ip of course). Can Steelapp repoints to another remote TIP, whenever local DC internet connection dies?

 

please advice. thanks

2 REPLIES
Occasional Contributor

Re: Disable VIP whenever WAN/Internet failure occurs

Ok all, just got some ideas and now testing using trafficscript rule at SMTP pool. Will share the script if all ok...

FYI, just the same person but had to change I'd as I forgot my password
Occasional Contributor

Re: Disable VIP whenever WAN/Internet failure occurs

HI guys,

 

just update on what i did...

 

1), create a pool called "Internet monitoring". Just fill any known/long lasting internet ip such as google dns (53), www.yahoo.com(80),4.2.2.2(53)etc, where port does not matter, as long as can ping. Then use "PING" as the health monitoring. Pls modify your necessary frequency to check your internet reachability.

 

2) Create the other data centre side pool, where the pool consist of a VS of the other side DC. I.e "Backup SMTP pool DC02", where the node is the other DC VirtualServer TIP.

 

3) In your "Main SMTP VirtualServer DC01", create a 'Request Rule', named "SMTP Redirrect", with one time check

 

4) Use the following traffic script;

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

if ((pool.activenodes("DC01 Internet Monitoring") == 0) && (pool.activenodes("Backup SMTP pool DC02") == 1)) {
   pool.use("Backup SMTP pool DC02");
   log.warn("Possible Internet failures, outbound SMTPs are routed to DC02 SMTP MTA");
}

pool.use ("Main SMTP pool DC01");

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

 

what it does, is that for every request made to your SMTP virtual server, it will check all your 'Internet Nodes' availability and your backup SMTP pool (the backup SMTP VS TIP), whereby all your SMTP traffic will be routed to your backup location if your current Steelapp cannot ping to all internet nodes and still sees your backup SMTP TIP or node IP alive.

 

if the "and" statement not valid, it will still use your main pool in DC01. This script is base on my network that have a mpls to connect both DCs and each DC has their own internet link.

 

you could always modify the script but the idea was to modify the path of SMTP traffic whenever local DC internet failure.

 

 

my thanks to the original trafficscript provider...