If you were watching Amy McCormick and myself on the MQ webinar on March 30th, you'll have had the fun of hearing me talk you through a demo of the new HA capability for MQ in containers. The good news is that the demo went exactly to plan. The bad news is that the screen sharing didn't! I've re-recorded the run through of what I did yesterday and uploaded it, just to prove I wasn't making it up ;-)



------------------------------
David Ware
STSM, IBM MQ Chief Architect
Hursley, UK
------------------------------

Thanks David,

I didn't get to the talk due to timezone issues, so seeing this was great.

I did have a question about the CPU usage behaviour that was visible during the demo.

For the first two or three minutes viewing the NativeHA pod resources, the first two pods had relatively high CPU (.2 to .3) compared with the idle single instance (< .03). The third pod usage was lower at 0.08 or 0.09. Then during the demo, the usage dropped back much closer to the single instance level.

Any thoughts on why we would see that sort of pattern?

Anyway... great demo and great capability. Can't wait for it to be production ready.

Regards,

------------------------------
Neil Casey
Senior Consultant
Syntegrity Solutions
Melbourne, Victoria
IBM Champion (Cloud) 2019-21
+61 (0) 414 615 334
------------------------------

Original Message:
Sent: Wed March 31, 2021 04:18 AM
From: David Ware
Subject: MQ Native HA demo in OpenShift

If you were watching Amy McCormick and myself on the MQ webinar on March 30th, you'll have had the fun of hearing me talk you through a demo of the new HA capability for MQ in containers. The good news is that the demo went exactly to plan. The bad news is that the screen sharing didn't! I've re-recorded the run through of what I did yesterday and uploaded it, just to prove I wasn't making it up ;-)

------------------------------
David Ware
STSM, IBM MQ Chief Architect
Hursley, UK
------------------------------

Hi Neil,

That's a good spot, but I'm not sure I know the answer. This may just be an OpenShift anomaly in when it's sampling. You'll see those numbers fluctuate, sometimes with all three in the 0.0x range.

The key thing is that the general MQ overhead is very low, which is something we often take for granted. Having such a low background overhead providing a critical messaging service is definitely not a given across the field.

Thanks

------------------------------
David Ware
STSM, IBM MQ Chief Architect
Hursley, UK
------------------------------

Original Message:
Sent: Wed April 07, 2021 03:51 AM
From: Neil Casey
Subject: MQ Native HA demo in OpenShift

Thanks David,

I didn't get to the talk due to timezone issues, so seeing this was great.

I did have a question about the CPU usage behaviour that was visible during the demo.

For the first two or three minutes viewing the NativeHA pod resources, the first two pods had relatively high CPU (.2 to .3) compared with the idle single instance (< .03). The third pod usage was lower at 0.08 or 0.09. Then during the demo, the usage dropped back much closer to the single instance level.

Any thoughts on why we would see that sort of pattern?

Anyway... great demo and great capability. Can't wait for it to be production ready.

Regards,

------------------------------
Neil Casey
Senior Consultant
Syntegrity Solutions
Melbourne, Victoria
IBM Champion (Cloud) 2019-21
+61 (0) 414 615 334

Original Message:
Sent: Wed March 31, 2021 04:18 AM
From: David Ware
Subject: MQ Native HA demo in OpenShift

If you were watching Amy McCormick and myself on the MQ webinar on March 30th, you'll have had the fun of hearing me talk you through a demo of the new HA capability for MQ in containers. The good news is that the demo went exactly to plan. The bad news is that the screen sharing didn't! I've re-recorded the run through of what I did yesterday and uploaded it, just to prove I wasn't making it up ;-)

------------------------------
David Ware
STSM, IBM MQ Chief Architect
Hursley, UK
------------------------------