Skip to Content
0

Classic Alert Framework not work in all cases

Feb 14, 2017 at 08:08 AM

21

avatar image

Hi all,

when we have setup our new PI 7.5 System we evaluated what is for us the better Alert approach. The new model where you have to subscribe for each interface or partner was at that time not suitable for us, also the fact that you have to schedule a e-mail job for each consumer. So we decided to use the old Alert Framework.

I have configured it like that for all partners (our partners starts always we a two letter ISO code) for sender and receiver direction, e.g.

I have tested it in the test system and it worked. Also on production, I have seen cases where I have received alerts but I also have plenty of cases where I have not received an alert. Is there a difference in the behavior with using wildcards between the steps or engines? If the error is channel related, it supposed to work where it looks like that in case the mapping has a problem the alert is not raised.

We are organized in regions, so based on the prefix of the partner we know who is responsible for the alert. And in each region we have a group of (sometimes) overlapping people.

My thoughts are:

- Find the bug and fix it.

- Going to new Alert Framework and implementing a report that is generating for each party an alert rule (Directory API, Alert API) and consuming the alerts with an own report that is also generating the e-mails

Any other thoughts from you?

Thanks
Helmut

untitled.png (8.7 kB)
10 |10000 characters needed characters left characters exceeded
* Please Login or Register to Answer, Follow or Comment.

0 Answers