We using BizTalk360 to manage the integrations for while now. We use the BizTalk360 application mainly for our Integration Management Groups. In each group there are one or more persons responsible for the integration flow.

Then there’s me, responsible for that the BizTalk plattform itself is running.

BizTalk360 are allowed to see message content on suspended instance, because they need to take action depending on content. This can be sensitive data. We already see that a person terminates or resumes a instance. What we lack is auditing when users are looking at the message content. By regulation we need to audit any views on sensitive data.

So we would like to make this a feature request to future versions of BizTalk360 that we can audit message content views, because auditing is a high priority for us.

By Johan Ljunggren

I have a big issue in the Dutch Bank about resubmitting messages from BizTalk 360. I have a http receive location that can be use to resubmit messages. It takes a long time to finally get it worked, but now anybody who knows the uri can submit an message in the message box this message can be a bank transaction to your own account :). So only BizTalk 360 may resubmit these messages to the uri. At this moment anybody with a wcf client can put a message in the messagebox because you can log in with anonymous. Is there a security way to solve this problem, because otherwise they can not use this functionallity and this was one of the reason they want to have BizTalk 360.

I want to protect the HTTP web service, for anyone who know the url for example the developers. Because he can put any message in the message box with anonymous login. Is there a way to protect this. The best way is to create a white list for user who can acces the HTTP web service. Another option is that BizTalk 360 is using default localhost port 80 to resubmit message through HTTP web service. Can I not change that to another port so we can block that port on the server so the only user who can resubmit must have acces to the server.

Greetings, Sergio

It is not possible to end a data process monitoring at 12h00am (midnight). Let say we want to monitor 7h00pm until 12h00am, the web interface got an error: "The end time should be atleast 15 mins later than the start time.". We have to set 23h45 or 00:15. So it is impossible to get a result at 12h00am.
If we set 23h45, we don't have the last 15 minutes of the day included in the monitoring. If we set 00h15, the first 15 minutes of the new day is part of the new day, not the last day.

So our results are not exact if we take a look in the data monitoring dashboard.

We have quite a lot of servers, and thus quite a few artifacts we monitor. However, we have way more artifacts (NTServices, SQL jobs, etc.) that we do not monitor. But I notice these artifacts still add rows to the table b360alertMonitor_LogStateCheck (with ExpectedState 2). This gives us a lot of useless rows in this table for things we don't want to monitor anyway. Is there a way to turn this off? We now have 17 million rows in this table, and we only keep this data for a week (the cleanup job works fine).
I can write a job myself to delete these rows with ExpectedState 2 at a higher frequency than 7 days (more like every hour), but I was hoping there was an option to actually indeed 'do not monitor' them at all

By Marc van de Klashorst

The requirement is that the customer wants to receive an alert each time a message or instance got suspended. So this is a bit like an event driven approach.
Applications
Say you have an alert for Suspended Instances with a Warning threshold of (more than) 0 and an Error threshold of (more than) 5. This would result in getting a notification when:
a. nothing is suspended yet and
b. a Message or Instance got suspended
This is because the Warning threshold is set to (more than) 0. As far as I am aware of, the next notification you would get is when the Error threshold is met. So when this is set to 5, you would not get notifications of suspended messages/instances 2, 3 and 4.
Of course you have the 'If violation persists...' functionality, but when the 2nd, 3rd or 4th instance(s) got suspended outside that timeframe, multiplied by 'Limit Number...', you would not get a notification for the instances which occur outside that timeframe.
Data Monitoring
With Data Monitoring you could set the Data Filter to Suspended Service Instances and the Warning threshold to > 0, but when there are no suspended instances within the configured frequency you would get an email that no instances got suspended, which is a bit over-informative .
As far as my experience with Data Monitoring reaches, the closest you can currently get is a daily report with the number of suspended instances of that day, even when that's zero...
To solve this you could think of adding an Information Threshold which can be assigned a number, just like Warning and Error Threshold under Applications. When assigned:
- 0 no notifications will be sent
- 1 a notification will be sent for each suspended instance
- 2 a notification will be sent for every 2nd suspended instance
- 3 a notification will be sent for every 3rd suspended instance
- etc.
To prevent flooding the 'If violation persists...' / 'Limit Number...' functionality would apply.

• Idea shared by Lex Hegt.

Would like to know how to monitor running services instances for duration > x minutes. I can see how to monitor by count, but not by duration.

We occasionally see the case where many different service instances are in a running state for an extended period of time (not by design). The issue generally turns out being at the platform level (ie. resource contention on the servers causing multiple hosts to throttle or just hang).

We will work on tuning the hosts, and addressing these types of issues. In the meantime, however, we would like to have awareness when any services instance, or a specific service instance, is running for longer than a certain period of time. Is this possible in BizTalk 360?

Again, I am looking for both options…

1) All services instances running for longer than x minutes generates an alert
2) Specific service instance names ‘xyz’ running for longer than x minutes generates an alert

By Syed Zaidi

I am configuring some Data Monitoring Alerts in BizTalk360 and have run into a snag. In my Test environment, we have crafted an ESB Exception for our Gibsons.FIN.Transactions BizTalk Application and have a Fault Category of ProcessCOACacheMisses. However when I am configuring this Data Monitoring alert in Production, I don’t have this Fault Category yet as it is a new application. I would imagine you are performing a DISTINCT SQL Query on the field and populating the dropdown with these values. It would be nice to override and provide my own text as now I have to wait for that alert in Production before I can configure the monitoring.

Not a huge deal as I will loosen my filters until the event happens but thought I would pass this feedback along.

By Kent Weare