I have two questions regarding alarm notifications.
Is there a way to make alarms that occur multiple times in a short timeframe not send multiple emails? For instance, if when a device is faulting, the condition goes high every 2 minutes, can I have Ignition only send 1 email per X minutes per alarm tag?
If I have an alarm that goes high and sends an alarm, and it is still high after Y minutes, can I have the system re-send the alarm email?
I know I could probably manage these things in Oracle by just logging the different alarm values I wish to notify on and sending emails out of the database itself, but it would be great if I could do this directly in the tag configuration.
Unfortunately, I don’t think either of these are currently handled very well. For #1, the closest we have is the time deadband, but that is slightly different- it means, “the alert condition must be true for this long before I report it”. If it flutters in and out and never stays high for that long, you’ll never see it.
For #2, the best you could do would be to set up your alarm as follows (oh, unfortunately this would only work in a transaction group, but you could set up a group that didn’t actually interact with the db): create an expression tag that boils your alarm condition down to a 0 to 1. Then back on the General tab, set it up as an Hour Meter, not retentive. Finally, set up the alert as an Analog alert with different levels for the time. You see, as an hour meter, the value will be time… so level 1 could be “active”, >0. Level 2 could be “Active for over 10 minutes”, when the value is >60.
If it makes you feel better, alerting is going to get a major overhaul in 7.3, and your #1 was one of the first things brought up in our first planning meeting.
Thanks for the reply. I figured as much, but wanted to check in to make sure. If necessary, I will handle it in a transaction group or in Oracle. I do think this type of functionality would be a valuable addition to the software, so I’m glad it is on the radar for 7.3 (I have a terrible phobia of generating massive amounts of annoying automated emails). I suppose while I’m at it, I’ll toss in a possible #3 to consider:
-Only generate an alert if condition goes high X times in Y minutes. This would allow the system to tolerate occasional brief spikes in data (with a duration less than the time deadband), but not repeatedly.
Hi everybody,
I have a similar question about point 1.
In my case, the system sends several emails at the same time when the alarm conditions is true.
How could I “limit” it in order that the pipeline send ONLY ONE email?¿
Thanks in advance,
Guillem