I’m curious what is the general best practice is considered for channel limits on a given Iguana 6.1 instance? I know there can be a lot of variability depending on the complexity of work a translator on a given channel needs to perform, but say I want to have 300 channels in a single Iguana instance that are relatively simple LLP pass-throughs with some basic message filtering, should I expect issues assuming it’s running on more than capable hardware? What are the guidelines? Should I segregate my more intensive channels to a dedicated instance that has say a limit of 50 channels?
I would say 100+ is the approximate limit. Issues come up with instances with 250+ channels but these are rough numbers. As you mentioned, it mainly depends on the type of processing.
Another consideration outside of hardware is maintenance, it becomes increasingly difficult to troubleshoot an Iguana with too many channels. For example, if a memory leak occurs and one channel is responsible, you would need to analyze each channel to find the culprit and this would be very difficult with 300 channels.
The biggest issue I saw with 380+ channels, was the Synchronous Writes to logging. Channel began backing up as it tried to keep up with all the writes. My setup is mostly LLP -> Process -> DB.
Disable at your own risk. Safer to split things up a bit I think.
Macro Helix / McKesson
Software Developer / Integration Lead / Scrum Master
You must be logged in to reply to this topic.