How many scheduled processes are considered "too many" ?

Let's assume that our Appian instance is going to manage several processes which are basically poller, doing these tasks:

  1. retrieve data from rest urls or datasources (the payload could vary between 1kb to 5mb)

  2. apply some kind of basic data manipulation on the retrieved data
  3. save the manipulated data in a database

Let's say that we have:

  1. 20 processes running 24/7 every 30 seconds
  2. 10 processes running 24/7 every 2 minutes
  3. 15 processes running 24/7 every 1 hour
  4. 10 processes running 24/7 every 6 hour

In conclusion, we have something like 55 processes, let's round them to 60, which are doing the tasks listed above. Is this setup critical in terms of performance and system reliability ?

  • Appian can handle a very high number of processes (there are Appian environments that run over a million processes per day); you just need to make sure that your design and configurations take your expected volume into account. 

    I'd recommend taking a look through The Appian Playbook  , particularly the sections on scaling and performance testing. You can also review the documentation page on scaling Appian .

  • It's not the number of processes that's going to be the problem, but the overall aggregate effect they're going to have on your system.  Fine tuning them will probably net you more performance boost than reducing the frequency with which they run.

    Check out how your database resources respond, what happens to your CPU and RAM.  The size each one gets to and the length of time they sit before being archived are going to be some big factors to look into.

  • As the others have mentioned, it is not truly the number, but instead their footprint. Keep your processes light and archive as soon as possible. Keep your processes light by reducing nodes, chaining, PVs, etc.

 Discussion posts and replies are publicly visible