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
11 replies
Subscribers
7 subscribers
Views
6852 views
Users
0 members are here
Share
More
Cancel
Related Discussions
Home
»
Discussions
»
Administration
I have a chained activity that takes a bit of time to get from one screen to the
jamesw353
over 11 years ago
I have a chained activity that takes a bit of time to get from one screen to the next. It's calling a web service, parsing data, and displaying everything in multiple paging grids. I've reduced the time it takes to run from several minutes down to about 30 seconds. The problem is that the form is closing and I have to go to the task list to pick it back up.
I changed the custom.properties timeout setting (conf.suite.REQUEST_TIMEOUT_IN_SEC=60) to 60 seconds, but it still closes after about 20 seconds. Is there another setting that needs to be changed to extend the timeout?...
OriginalPostID-67771
OriginalPostID-67771
Discussion posts and replies are publicly visible
0
sonias
over 11 years ago
Hi, Sometimes while re-mapping your nodes, there is every chance that the chain is broken somewhere. Say inside a sub process that gets called in the chain. I am sure this is something you would have already confirmed. A quick test would be to remove the node that consumes the most time from the chain and check if timeout is your real problem.
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel
0
Jorge Sanchez
Appian Employee
over 11 years ago
Also, make sure that you check the "Allow chaining into the subprocess" if you are using subprocesses.
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel
0
shelzle
over 11 years ago
In these cases I build a chained parallel process flow that shows a "Please wait" form to the user which auto-submits after a few seconds and loops until the longer lasting flow has finished. Make shure that the process handles closing of the window and the case that the flow does not finish.
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel
0
jamesw353
over 11 years ago
I'm still looking into this. I triple checked all the activity chaining and everything looks good. I also tried removed all the gateways. Nothing seems to be helping though. I have two script tasks that run multiple instances of themselves. One is in sequence the other in parallel, which both run in parallel by forking the process flow with AND gates on each end. Could executing the scripts multiple times cause an issue? The documentation mentioned something about running multiple chained form instances, but doesn't warn of any issues with multiple chained script nodes.
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel
0
sonias
over 11 years ago
Can you please check number of times MNI running on the script tasks exceed the threshold value. If it exceeds, there could be a chance of breakage in activity chaining. Hope below link will help you:
forum.appian.com/.../Configuring_Activity-Chaining
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel
0
jamesw353
over 11 years ago
That was it! I read that page before but must have skipped over the line about multiple instances being included in the count. I was at about 20 for total individual nodes, then one of the multiple instances was running about 40 times in this test file. Increasing the max number chained of nodes fixed it. Although, I may end up using a waiting page instead. The number of instances to run is dynamic and could cause an issue. Is there a drawback to increasing the chained node limit that you're aware of? I'm keeping a close eye on server performance just in case. Thanks for the help!
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel
0
bruceh
over 11 years ago
We have increased the node limit to 80 and have not experienced any issues.
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel
0
Jason
Appian Employee
over 11 years ago
While raising the limit will prevent the chain from breaking, you may still want to evaluate whether it is necessary to run multiple-node-instances. If you could replace the MNI with looping functions, for instance, you would see a performance improvement in your wizard.
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel
0
bruceh
over 11 years ago
Our use case differs from James's in that we do not use MNI but we utilize many DB querys to retrieve data. We try to combine our querys as much as possible but we always seem to run into really complex SQL statements.
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel
0
Bishnu Panigrahi
Appian Employee
over 11 years ago
As Jason mentioned, replacing MNI with looping functions can possibly get rid of the need to raise the chaining limit, and therefore improve performance. Check out the comments on these posts to see the downside of increasing this limit:
forum.appian.com/.../b-45057
forum.appian.com/.../b-55644
Cancel
Vote Up
0
Vote Down
Sign in to reply
Verify Answer
Cancel
>