Comments on: Aggregated monitoring of BizTalk solutions using "BizMon"

By: Bart

Bart — Tue, 14 Apr 2009 14:48:32 +0000

Great Post, looks like a very very good tool!
Is there any more activity on it?

Hopefully you’ll get over the legal issues….

By: Arihant Jain

Arihant Jain — Fri, 27 Feb 2009 03:00:24 +0000

That really cool, let me know when you are releasing its beta so that i can try to use it in my project and and evaluate it and suggest more inputs to you.

By: Richard

Richard — Fri, 13 Feb 2009 08:27:02 +0000

@Mikael Håkansson: First I actually thought about releasing it without any support for setting up TPE-profiles and that users would have to use TPE-tool for that. But as I’ve seen my internal users struggle with this I’ve decided to at least look into using the approach you have in your code and see if I can support this within the actual tool somehow.

That would also make it easier to actually create one database model per application and not have on big activity for all. I think one big activity for all application wouldn’t really scale that well performance-wise one we use it some of our mote traffic intensive integrations.

@Arihant Jain: Yes, I’ve just finished the alerts functionality. It’ll be in there. You can set up subscriptions and have a list of email connected to that subscription. A subscription can then monitor x number of applications. Email is sent out as soon as something fails or a message get stuck. Another email is sent out once the subscriptions doesn’t contain any errors anymore. An NT Service is used to the checking and sending of email.

By: Arihant Jain

Arihant Jain — Thu, 12 Feb 2009 21:39:56 +0000

Hi the tools look great and i would love to contribute if required, our project needs such tool cuz its very difficult to go everytime on prod box to see the biztalk admin console for health monitoring and if such tool is available we can access it from our own system, i would suggest to include some alert functionality.

By: Mikael Håkansson

Mikael Håkansson — Fri, 23 Jan 2009 21:35:48 +0000

Looks really good, and I’m curious to see your generic activity, although I’ve your related posts. Are you providing any tooling for enabling tracking for ports or are you relying on TPE?
Looking forward to see you post it on CodePlex!

//Mikael

By: Richard

Richard — Fri, 09 Jan 2009 10:27:08 +0000

@Santosh

Yep, all the aggregation, status checking etc will be done by a separate project and those dll:s will be GAC:ed. These assemblies will be used both by the web GUI layer (as shown in the post) but also by a windows service handling sending all the alerts etc and could of course be used by other clients as well!

By: Santosh Benjamin

Santosh Benjamin — Fri, 09 Jan 2009 09:56:44 +0000

Richard,
Just came across this post and need to go through in more detail, but this is a great idea.
I think CodePlex is a good place to host this.
Regarding the architecture, if you adopt (or have adopted) a model where all the information aggregation is done by UI independent components (and even available through say, webservices) , then as Johan suggested, it could even be plugged into the ESB Portal or other portals (including say WPF clients).

I’ll send more feedback soon.

Regards,
Benjy

By: Henrik Møller Jørgensen

Henrik Møller Jørgensen — Thu, 08 Jan 2009 12:13:28 +0000

Richard,

Just joining the chorus here, and cheering for a release. I’ll add to the code if I have the skills required.

Henrik

By: Chris Leger

Chris Leger — Wed, 07 Jan 2009 15:58:09 +0000

I can definitely see the usefulness of something like this. HAT is flat-out miserable to use. Admin Console can be sluggish at times, and failed message routing in 2006 is better, but not ideal. We have implemented ESB-like cross-system consolidated exception logging, so there is yet another source of diagnostic info we look at.
So jumping from place to place, system to system to diagnose a problem is one of those tasks that one takes for granted because that is what we have been given by MS, and “that’s how it has always been”. I’m always open to 3rd party solutions that can make life easier. (case in point – I am a big fan of the BizTalk Deployment Framework on Codeplex)
I’m anxious to see what comes out of this initiative. Good luck.

By: Richard

Richard — Wed, 07 Jan 2009 08:29:38 +0000

@Johan Hedberg: Thanks!

I’ll try and publish a post discussing the architecture shortly but yes you’re right, all the history is based on a generic BAM tracking model. One of my goals is to make this as easy as possible to configure. I’ve worked with similar solution requiring adding pipeline components to all my pipelines etc and that’s not working for me. This should act as an extra layer on top of your existing solution – just as BAM works. But as you know BAM requires a lot of configuration so I somehow have to make that easier but that definitely has lower priority and will be something I’ll try to get in there in later versions.

When it comes to adding extra error information from suspended messages etc I’m not sure. My idea is to create a tool that aggregates information an could tell support persons that “something is wrong and someone need to look at this”. For all the detailed information we have the existing tools … I don’t want to replace something just aggregate and collect the information into a dashboard. But, that’s just my idea and could change (hey, this is open-source )! Love the input!

I haven’t looked that much into the ESB Guidance portal (just the exception management) but it looks interesting. All this is however build on ASP.NET MVC, jQuery, LINQ and I don’t think it’d be that easy to merge the two. It could be interesting to look into further down the road.

What’s important to me is that we create something that could exist on top on any BizTalk solution, based on ESB or not and that it’s dead simple to configure and setup.

I’ll get back to you as soon as I get a stable first version up on codeplex!