I want to build a threshold so that when the number of stations registered in a region drops by say 10% or more than 10 handsets, we get an email alert.
The individual station registration information is available in AVayaIPStationRegisteredIPStations, and I can build a display which shows me the information by region. I can write a threshold when an individual station is not registered but I can't figure out how to write a threshold that will alert us when the total number of stations registered drops below a threshold. I assume I need to build some sort of summary that I can then monitor with a threshold.
Anyone done something similar or who can point me in the right direction
unfortunately, combines are not supported in the thresholds. In order to achieve what you mentioned, we either need to create a new summary record, or do it using an analyst. If you needed it available in a record so that you can create custom thresholds on it, check with your account manager for further assistance.
The analyst for this is VERY easy and something we typically implement with our Automated Teller Machine (ATM) monitoring solution... I'm not sure what record you need this for but it would not take me very long to adapter it to any record that has a status to trigger off of..
Are you familar with Analysts or have any service hours?
I haven't looked at Analysts before but interested to see how this could be done. IS there some documentation or tutorial you could point me to, there doesn't seem to be much in the help files
I don't think we have any service hours
The Prognosis online help has a section about Analyst to start with. Please visit the links below for details.
Writing Analyst Rules:
It also provides many examples:
Using Analyst function, you could use multiple Prognosis records (unlike threshold condition in which you can only use one record in one condition) and do data calculations too based on the where clause or IF condition, for example. For this particular requirement, you can calculate the Total of registered stations and send alert.
If you need further assistance in building Analyst rules please engage our professional service team to help.
Our Help on Analysts is there but it hasn't been updated/gone through in a long time.. so it is incredibly unorganized, hard to follow, and doesn't follow any best practices.. the examples that are there also don't supply what you want but it is at least an example. .
If you tell me the record when I get back from traveling on the road I might be able to post a full example doing what you want.
We offer a training course and if you attended summit there is the deck I used there.
There are a couple of places the data is available AvayaPhone and AvayaIPStationsRegisteredIPStations.
My preference is to use AvayaPhone as that has other phone objects I would like be aleerted to if they were down.
Screen shots of the two data views.
Sorry the delay..been pretty busy..
This example is part of a lab example I use on a simple count exercise during the Analyst training when I deliver it.
What it does is uses a global variable to count track from interval-to-interval the number of down ATMs.. this one checks every 15 minutes.
At a high level.. this is a very basic skeleton example and you can do a lot more.
1) Every 15 minutes it initializes the local variable to zero for the current interval count.
2) It counts the number of ATM terminals in the secondary rule (which is where you would put your Avaya Record name) the primary rule calls.
3) It compares it to the previous interval and alerts if the change is greater than 50.. but you can modify this criteria.
4) It logs to the PROBLEM record.
In the lab I like to ask people to then modify the criteria and then also add additional destinations like email, SNMP, etc. I know for customers I have built ones that have Analysts invoke PQL (11.5+ required) and do some more advanced functions as well.. and do percentage changes or change over multiple intervals.
If you need for us to build something more specific I need to refer you to your account manager, but otherwise I would suggest giving it a try yourself.. as this should work at a high level for you.
SECTION CONFIG MAX_SIZE 5.00 MB MAX_DAYS 14 NUMERIC _LstCnt_g := 0 END_SECTION SECTION RULE_DEF RULE P001_BeginCheck PRIMARY NUMERIC _Count RECORD PNODES REFRESH 15 MINUTES ! Testing, comment me! ! REFRESH 10 SECONDS ! EVERY 24 HOURS ACTION SET _Count := 0 START RULE S002_CheckATMs WAIT END_ACTION ACTION WHERE (_LstCnt_g <> 0) and (_Count > _LstCnt_g) and (_Count - _LstCnt_g > 50) LOG 0001 IMMEDIATELY END_ACTION ACTION SET _LstCnt_g := _Count END_ACTION END_RULE RULE S002_CheckATMs SECONDARY RECORD ATMMON WHERE STATUS = "DOWN" REFRESH 10 SECONDS ACTION SET _Count := _Count + 1 END_ACTION END_RULE END_SECTION SECTION MSG_TEXT MSG 0001 "0001:Many ATMs down, @_Count@ currently down!" END_SECTION
Another thought on this one...
Could you not update Prognosis to track output of 'status socket' for any fluctuation ?
Unsure if this command is used on the product, however.
Total Registered IP Endpoints: 6584
Registered IP Endpoints with TCP Signaling Socket Established: 6575
Total Socket Usage: 06632/19660
Yes we probably could I don't see why not. I'd need to look into whether it is specifically used...but three short term courses of action since changing any part of Prognosis for additional functionality usually is fast.
1) Submit a RFE request with your Account Manager requesting these stats specifically be added to Prognosis.
2) Engage Services to do it for you now as an Extended Platform Solution; I am not so much in the UC Collaborate space of our product but this seems like something we could turn out from services VERY quickly.
3) If you know Prognosis Extractor and have the ADI license and available user records you can build the extractor yourself.
If you reach out to your Account Manager on #1 or #2 refer them to me and I can clarify for them how I'd go about this.