We are looking at upgrading our environment to add in HA paris for every poller and AWS system. It became obvious to me that we don't actually know how a update activity with HA would work, or what it would look like. I get how the HA functions with unplanned service outages, but here is what i'm confused on: Are we able to update the offline B set, then switch to the A set and preform the upgrade there? How would that work with a Database? Wouldn't there be a period of mismatch between the DB and the Pollers? Any pointers to documentation or better yet examples would be greatly appreciated, and Anyones hands on experience would be just Fantastic. Thanks all KMSigma.SWI aLTeReGo cause they knows all.

Orion's High Availability does not provide for zero-downtime upgrades of the Orion platform itself. It provides high availability from network, server, and application failures, as well as allows you the ability to patch or upgrade the operating system of another member of the pool without disrupting services. It can also be used to provide zero downtime migrations to new hardware or VMs. Upgrading Orion product modules, however, still requires some amount of downtime. Though that has been significantly reduced in the latest 2020.2 release.

This discussion has been locked. The information referenced herein may be inaccurate due to age, software updates, or external references.

You can no longer post new replies to this discussion. If you have a similar question you can start a new discussion in this forum.

New to HA, How Do Updates Work?

rfackrell over 3 years ago

We are looking at upgrading our environment to add in HA paris for every poller and AWS system.
It became obvious to me that we don't actually know how a update activity with HA would work, or what it would look like.
I get how the HA functions with unplanned service outages, but here is what i'm confused on:

Are we able to update the offline B set, then switch to the A set and preform the upgrade there?
How would that work with a Database? Wouldn't there be a period of mismatch between the DB and the Pollers?

Any pointers to documentation or better yet examples would be greatly appreciated, and Anyones hands on experience would be just Fantastic.
Thanks all

KMSigma.SWI aLTeReGo cause they knows all.

Top Replies

aLTeReGo over 3 years ago +1 verified

Orion's High Availability does not provide for zero-downtime upgrades of the Orion platform itself. It provides high availability from network, server, and application failures, as well as allows you the…

Parents

0 aLTeReGo over 3 years ago

Orion's High Availability does not provide for zero-downtime upgrades of the Orion platform itself. It provides high availability from network, server, and application failures, as well as allows you the ability to patch or upgrade the operating system of another member of the pool without disrupting services. It can also be used to provide zero downtime migrations to new hardware or VMs. Upgrading Orion product modules, however, still requires some amount of downtime. Though that has been significantly reduced in the latest 2020.2 release.
Cancel
Vote Up +1 Vote Down

Cancel

Reply

0 aLTeReGo over 3 years ago

Orion's High Availability does not provide for zero-downtime upgrades of the Orion platform itself. It provides high availability from network, server, and application failures, as well as allows you the ability to patch or upgrade the operating system of another member of the pool without disrupting services. It can also be used to provide zero downtime migrations to new hardware or VMs. Upgrading Orion product modules, however, still requires some amount of downtime. Though that has been significantly reduced in the latest 2020.2 release.
Cancel
Vote Up +1 Vote Down

Cancel

Children

0 rfackrell over 3 years ago in reply to aLTeReGo

This is a great explanation and general outline. I'm sure it's paperwork outlines this too, but this makes it plain. Thanks!
Cancel
Vote Up 0 Vote Down

Cancel
0 bobmarley over 3 years ago in reply to aLTeReGo

This is a fantastic and most timely response for a situation I am currently facing. I am proposing to our management that we spin up an exact duplicate at our second data center and work on the automation to keep the nodes, reports, and alerts replicated from the primary to the secondary. That way we can flip our NOC over to the secondary server during downtimes so essentially the system never goes offline during upgrades.
Cancel
Vote Up 0 Vote Down

Cancel