Unsolved
This post is more than 5 years old
6 Posts
0
1269
Watch4Net: Collector Resilience
Hi All,
Hope this is the right support forum to post a W4N related question, if not will move it somewhere more appropriate. I have two questions on Collector Resilience in W4N.
1. What is the standard implementation out there for resilience of collectors? Our idea is two collectors collecting the same data going into one Backend, and either just have both sets of data in the backend (taking an average of the two feeds for example), or have one collector live and one in standby, with the standby coming up if live goes down, or have the backend disregarding one of the collectors data.
2. In the above ideas, how would we implement any of these? For instance, how would we tell the backend to disregard the data from one of the collectors?
Thanks for any help.
AlanG2
17 Posts
1
July 25th, 2014 07:00
You run both collectors hot and feed into an instance of Crossed-Failover-Filter on the backend. This filter selects one source as "active" and discards all values coming from the other source. If the active source fails to send data for a pre-configured period of time then the filter switches to using the other source.
I tested this is the lab and it all seemed to work ok, but never actaully used it in production.
Alan
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
February 2nd, 2017 15:00
You can also review the high availability and Failover document available from support@emc.com and by clicking the following link:
https://support.emc.com/docu78976_High-Availability-and-Failover-for-EMC-MANDR-6.7:-Architecture-and-Use-Cases.pdf?langu…