Recently I have been testing redundancy across two containerized instances of Ignition v8.1.21 and have noticed that memory tags in all of my tag providers are losing value when the redundant server takes over. I was under the impression that the redundant server would pull updates for the tag providers as needed - I have even upped the Standby Activity Level config value to "Warm" to try and ensure this behavior. On the redundant gateway, the status page for Redundancy shows that we are pulling the updates from the tag provider pretty regularly. Despite this, the tags seem to be still losing value on failover. Interestingly, I am also losing configuration changes to tags, such as enabling History; once the master takes back control, a tag that had history enabled on it before the failover test no longer has history enabled.
Can someone shed light on this? Is this the expected behavior of tag providers when failover occurs, particularly with respect to the values of memory tags? I am (baselessly) wondering if there is something amiss with my containerization and dev environment, as we have some quality of life scripts that setup most of the environment for new members of the team, and it seems like the tags in the tag providers have values that line up with an import of our most recent exports.