| Author |
Message
|
| RA_SH |
 Posted: Tue Jul 17, 2012 12:56 am Post subject: Broker Health Check |
 |
|
Novice
Joined: 17 Jul 2012
Posts: 13
|
Hi All,
I am required to do a general health check for the Broker and the flows associated along with the execution groups.
Could someone please point me in the right direction as to, what are the points i need to keep in mind that would constitute a health check,
I could think of:
1. Broker status
2. Execution group status
3. Flow status
4. space availability
5. CPU Usage
Any other inputs, and how can i go about achieving these.
Any documentation that is available with these information about performing health checks on broker and resources.
Thanks |
|
| Back to top |
|
 |
| sourdas2 |
| Posted: Tue Jul 17, 2012 1:32 am Post subject: |
|
|
Voyager
Joined: 21 Apr 2006
Posts: 90
Location: Kolkata,India
|
|
| Back to top |
|
|
| RA_SH |
| Posted: Tue Jul 17, 2012 1:56 am Post subject: |
|
|
Novice
Joined: 17 Jul 2012
Posts: 13
|
Thanks!!! but i have already checked this. Are these the only parameters...???? for the health check....
If so, then this covers the broker resources health check..
Any pointers for the deployed artifacts health check...??? |
|
| Back to top |
|
|
| mqjeff |
| Posted: Tue Jul 17, 2012 2:57 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
You should revisit the contract you're on that has requested that you perform this health check.
There is no industry defined standard for "health check" for software in general, much less for Broker in specific.
So you're really in a customer-specific situation, where you need to make sure you understand what the customer expects, and make sure you understand what you've been hired to do. |
|
| Back to top |
|
|
| RA_SH |
| Posted: Tue Jul 17, 2012 4:14 am Post subject: |
|
|
Novice
Joined: 17 Jul 2012
Posts: 13
|
Hey mqjeff...
This is not any customer/client specific requirement, its just an internal thing, some sort of an enhancement thingi that i was thinking about |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Tue Jul 17, 2012 4:23 am Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
What applies to one site will not always apply to another.
For example, one check could be
Are there any message on any Backout Queues?
That won't mean mush to a site where all the traffic is SOAP over HTTP.
Then you have to ask yourself this question
How far do I go to verify that my broker is running sweetly?
It could be that just seeing regular traffic via Event or Flow monitoring is sufficient to answer 'yes'.
then you might decide to connect up a tool such as Tivoli or HP-Openview and let it do all the work for me.
Enjoy!
_________________
WMQ User since 1999
MQSI/WBI/WMB/'Thingy' User since 2002
Linux user since 1995
Every time you reinvent the wheel the more square it gets (anon). If in doubt think and investigate before you ask silly questions. |
|
| Back to top |
|
|
| lancelotlinc |
| Posted: Tue Jul 17, 2012 5:00 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
JMX is available for the JVM that the Broker runs. You can collect stats from it as these stats relatewithin the JVM.
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
| mqjeff |
| Posted: Wed Jul 18, 2012 7:57 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
| RA_SH wrote: |
Hey mqjeff...
This is not any customer/client specific requirement, its just an internal thing, some sort of an enhancement thingi that i was thinking about |
Define the questions you want answered about your environment.
Or define the decisions you need to make over the next year - if only in category if not in specifics. For example "will we need to move to larger hardware?" "When will we upgrade to the next version of the product".
Design your healthcheck to support those decisions. |
|
| Back to top |
|
|
| RA_SH |
| Posted: Wed Jul 18, 2012 9:31 pm Post subject: |
|
|
Novice
Joined: 17 Jul 2012
Posts: 13
|
Hi All,
After a lot of discussions and a brainstorming session, we have concluded that the Generic Health Check would basically consist of three stage check.
1> A system health check, where we can have scripts to monitor the disk space usage and CPU usage and alert the admins if any abnormalities.
2> Deployed resources check, this would include, checking the status of the Broker, QM, Exec groups, Flows and depths of request queues.
3> Services check, where we would have scripts to fire webservice or MQ requests for each of the service and check the responses we get. The request messages would be read from a property file.
What do you guys think of this approach.
I will still keep troubling you guys with how to go about achieving this..  |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Wed Jul 18, 2012 10:19 pm Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
| RA_SH wrote: |
3> Services check, where we would have scripts to fire webservice or MQ requests for each of the service and check the responses we get. The request messages would be read from a property file.
|
Naturally this will put a loading on the services in question and could affect user response time.
From my own experience here unless you are very careful you could actually make your system worse. Think very carefully about the type of requests you are going to send. Can the target flow handle this sort of request or will lthey need modification? you really don't want to make a time critical process slow down with silly H/C requests.
If you go beyone simple 'keep alive' or 'ping' type requests then you will go beyond the basic H/C operation and into the 'continuous UAT' type of environment.
A very different scenario and one that is usually more trouble than it is worth.
_________________
WMQ User since 1999
MQSI/WBI/WMB/'Thingy' User since 2002
Linux user since 1995
Every time you reinvent the wheel the more square it gets (anon). If in doubt think and investigate before you ask silly questions. |
|
| Back to top |
|
|
| RA_SH |
| Posted: Wed Jul 18, 2012 11:32 pm Post subject: |
|
|
Novice
Joined: 17 Jul 2012
Posts: 13
|
| The requests will be simple ping request, just to check if the flow is running fine. |
|
| Back to top |
|
|
| Vitor |
| Posted: Thu Jul 19, 2012 4:57 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| RA_SH wrote: |
| What do you guys think of this approach. |
Nice wheel you're reinventing there.
| RA_SH wrote: |
1> A system health check, where we can have scripts to monitor the disk space usage and CPU usage and alert the admins if any abnormalities.
2> Deployed resources check, this would include, checking the status of the Broker, QM, Exec groups, Flows and depths of request queues. |
Whatever product your platform people are using to check the health of the OS & underlying hardware will already be doing 1). There's absolutely no point building something for this as the TCO simply will not justify it & I'll be amazed if your platform people (even if they have built something) don't have something that's doing this. Any commercial product you use either do 2) for you or be easily extended to do it for you. Tivoli & Openview have been mentioned but the facility is shared by all products
| RA_SH wrote: |
| 3> Services check, where we would have scripts to fire webservice or MQ requests for each of the service and check the responses we get. The request messages would be read from a property file. |
What exactly is the point of this? If you actually need to check that the responses to a request are correct on a regular basis, then there's something seriously wrong with your change control procedure. You do not need to do a regression test with messages every 15 minutes to see if the flow's working. You might well want to see how fast it's working, and the WMB facilities will tell you that without sending data into your production system. What "simple ping request" were you planning to use to test that a flow's working which updates something?
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
|
|
