View the latest IBM Application Performance Management questions on our forums and see the answers provided…

IBM Performance Management forum – Week of May 8th

overhead the agents on system

Hi,

Can you please inform which affect on system doing agents? (CPU/memory/and other)

Interesting the folowing agents:

a) NET agent

b) IIS-agent

c) Responce time monitoring agent


SaaS: New WAS version installed on new server. Same host but different IP. Can we change the agents instance name?

Hi, We are currently preparing to migrate WAS and IHS to later version and these are installed on new pre-prod servers while the existing old WAS/IHS are running in parallel.

These new pre-prod servers’ hostnames are exactly the same as the prod servers except that they have different IP addresses.

While attempting to install the Linux, WAS and IHS agents on this new pre-prod server, we noticed that the instance name of the agent e.g. host1:LZ was automatically set and that conflicts with the existing prod server setup.

So in the GUI, when I click on the existing host1:LZ, it is pointing to the new pre-prod server instead of the prod server.

Question:

1. Is there anyway to define the agents’ instance name during installation? e.g. instead of auto-picking up the hostname and name it {hostname}:LZ, can I define a new name e.g. {hostname-new}:LZ? I suspect the same problem will appear for the rest of the WAS and IHS agents.

2. Or is there a better way for handling such ‘migration’ e.g. create a APM SaaS trial instance, have everything configured the same as the existing prod envt, and then upon the cutover day, update global.environment file to point to the APM SaaS production instance instead? Is this the only file that needs to be updated?

Option 2 would have been a preferred option though as we want to retain the hostname as the instance name during the cutover. Any advice is greatly appreciated. Thanks! Nicole


Monitoring MQ File Transfer agents

Is anyone monitoring Websphere MQ File Transfer agents using IBM Performance Management? How are you doing it?


APM Synthetic Transaction SubTramsaction not sent in EIF Alarm

Hello, We are doing a transaction every 5 minutes against a full selenium test suite. Everything is cool. We also did configure a threshold that if the PGOOD (Percent Good) is less than 100% 2 times in a row, it does send an alarm to OMNIbus through eif.

But our problem is that we only have the name of the whole transaction, we do not have the failing(s) subtransaction(s) in the eif alarm. This is quite a problem since our support team receive the alarm that “there is something that failed in transaction”, but without details of what is failing (which portion of transaction).

We would need that all the subtransactions status would be sent in the eif alarm, so we could attach that to the OMNIbus alarm. This is obviously not the case.

One solution would be to use Impact to “enrich” the transaction failed alarm by getting the subtransactions list from APM rest interface, but there is no indication on how we could do this.
Any of you have any idea of how to do this? Thanks


dwAnswers – Week of May 8th

APM 8.1.3: DB2 levels

What are the DB2 levels that are supported by APM8.1.3?


Join The Discussion

Your email address will not be published. Required fields are marked *