Appian Community
Site
Search
Sign In/Register
Site
Search
User
DISCUSS
LEARN
SUCCESS
SUPPORT
Documentation
AppMarket
More
Cancel
I'm looking for ...
State
Not Answered
Replies
4 replies
Subscribers
8 subscribers
Views
2316 views
Users
0 members are here
Share
More
Cancel
Related Discussions
Home
»
Discussions
»
Administration
High Availability Active-Active Setup
andrewl657
over 8 years ago
Hello,
I'm looking for more information of why Appian suggests Active-Passive setup vs Active-Active for High Availability. Our organization has to have Active-Active for our current setup but we are being told Appian 16.2 has an issue with Active-Active and it won't be fixed until the next scheduled release. Can you please give more information into what the problem is and how we monitor? Also if fixed in next release when will that be?
Thanks.
OriginalPostID-230918
Discussion posts and replies are publicly visible
0
juergeng
over 8 years ago
What I know about that issue is, that Active-Active needs a Network response < 50 ms between the involved systems (and if >50ms the fall-over procedure will take place - I guess with possible swapping back and forth). The next release will be 16.3 if this is meant to be the next scheduled release. This should be taken place in the next couple of weeks - at least that is what the name intents to say: 3rd Quarter.
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel
0
andrewl657
over 8 years ago
So the response time of 50 ms is the responsibility of our network to have continuous communication so data files do not become corrupt(out-of-sync)? I imagine this could potentially be high downtime. Is anyone else using Active-Active and how much are you encountering these issues? Any tips to minimize down-time, file corruption, etc.?
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel
0
juergeng
over 8 years ago
Yes - the 50 ms is the max. response time of your network for the heart beat of the K-engines (Appian back-end engines - running as separate OS processes). In case of failover, the Backup Engines will step in to perform the necessary work. This demands that both Primary and Secondary Engines (Back-up Engines) share the same KDB-Files. It has nothing to do with Business Databases. If the Platform gets busy with handing over the work from Primary Engines to Secondary Engines because it thinks it is necessary (lag time of network) than it can be swapping back and forth and only less computing resources are left to respond properly to process execution. For more information please see:
forum.appian.com/.../Hardware_Engine_Failover_Configuration.html
and
forum.appian.com/.../High_Availability_and_Disaster_Recovery_Configurations.html
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel
0
andrewl657
over 8 years ago
Great. Thanks for the explanation!
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel