Hello,
I’ve discovered the wm.tn.route:load, which refreshes TN information for processing rules.
This leads me to two questions:
the wm.tn.route:load service reloads the processing rules, but what happens if a document arrives at the same time? Will it wait for the load to finish, will it use the PRs as they were before the load? (the real question is: can I run this service on a production environment without messing up the routing?)
We often have requests for modifications of the processing rules (like "my client now sends files with filename toto.csv instead of titi.csv).
Up to now we modified the processing rule in the TN console, and sometimes the new parameters weren’t efficient until the next reload of the package WmTN on each IS that is linked to this TN database. The problem is that reloading WmTN can take several minutes.
I recently discovered the “load” service and am currently investigating more on its advantages and side-effects.
By the way: is there a documentation for the WmRoot, WmTN, and other packages besides WmPublic? I think there are a lot of services there that might be really handy, but I just don’t know what they do
Thanks in advance
PS: I’m working on IS 6.1, TN 6.1, soon to become 6.5 for both.
TN can be configured to notify other instances that the DB has changed and thus they should reload their caches. Refer to the TN docs and the TN properties admin help page for information on setting this up.
Modifying processing rules on the fly in production is risky, IMO. One mistake can wreak havoc on the system. Undoubtedly you’re not in a position to do this now, but over time it would seem a good idea to structure the rules such that they are immune to things like filename changes and such.
WmRoot is not documented. It is not intended for our use. WmTN is documented in the TN Reference. Services that are not documented are not intended for our use.
Cool I found the solution in the TN Docs.
Indeed it seems cleaner to do it that way!
Concerning the change in processing rules, I totally agree that it is a bad habit to change it in the middle of a working day, but it is not my decision to make
Besides it can sometimes also be important to block an integration service for a specific file, and as far as I know, the simplest way to do that is to disable the processing rule, isn’t it?
Thanks a lot, there seem to be a lot of interesting features in the documentation we know nothing about, maybe we should take a read at it someday :p:
BTW: the solution was p.48 in the TradingNetworksUsersGuide.pdf (Configuring Trading Networks for a Clustered Environment)
That depends completely on the nature of the integration. By disabling the rule, one of two things will happen: 1) the document will be processed by another rule; 2) the document will be completely ignored and never processed again (without manual intervention). Depending on the integration, either of those options may be quite undesirable.
In other words, the document isn’t “blocked” which implies that enabling the rule later will then process the document. The document will be processed in some manner, perhaps ignored, which may not be what you want to have happen.
