What is a CVO Freeze event Not Before time
Applies to
Answer
ONTAP constantly polls for freeze events from Azure.
When it receives a response that one is about to occur, it will begin the graceful Takeover (TO) and reboot of that node.
There are several types of freeze events that can occur, and these tend to occur outside of the normal scheduled Azure maintenance events. Many of them are benign, but one of them can compromise the integrity of your data. Since we are not yet able to identify what kind of freeze event is occurring, we have to treat every freeze event like it is one that will compromise your data.
Therefore, anytime a freeze event occurs, ONTAP will gracefully takeover and reboot the node that is being affected by a freeze event.
The node goes into TO right away so we can prepare for the event... that is we need to be not serving data, we've done a reboot, and are waiting for the event to complete. So, we can't use the 'Not before' time as any type of guidance for the storage admin to tell their folks that they have x amount of time to prepare for this maintenance event.
When we get an alert that event a freeze event is going to occur itt has an ID of xxxx-xxxx-xxxx-xxxxx.
Wed Dec 11 06:19:07 -0500 [Node-01: mgwd: vsa.scheduledEvent.scheduled:alert]: Cloud provider event scheduled from: Platform, out: 6 for node "Node-01".
ID: 9BFB6755-ACD6-4B1C-9596-xxxxxxxx.
Type: freeze. Not before: 12/11/2024 06:33:33.
The event won't happen until 12/11/2024 06:33:33.
ONTAP, because we don't know what type of event, goes into TO, and then reboots to a certain point, waiting for the event to complete.
Once that event completes, we reboot again so we start from a clean state.
so the 'Not before' time is when the Azure maintenance event is going to occur... not when the TO begins