C9300X stacks have been reported to fail due to memory exhaustion

[Ajesh]

Hi Team,

I am planning to upgrade the firmware on our Meraki Switch C9300, and while reviewing the release notes for the latest firmware version, I came across the following known issue:

C9300 known issues
Some C9300X stacks have been reported to fail due to memory exhaustion, requiring an IOS restart to restore service.

I would like to confirm whether this issue is known to occur only during the firmware upgrade process, or if it can also happen during normal operation.

Appreciate your guidance on this before proceeding with the upgrade.

[alemabrahao]

This issue can also occur during normal operation, not just during or immediately after a firmware upgrade.

[Ajesh]

Could you please confirm whether the memory exhaustion issue affecting C9300X stacks is present only in the latest firmware version (17.2.1.1) or if it was also observed in previous versions?

Additionally, if the issue occurs during normal operations (not just during upgrades), would proceeding with the upgrade to 17.2.1.1 still be considered a feasible and stable solution?

[alemabrahao]

Checking the release notes of other versions, I only noticed this problem in version 17.2.1 and the current one.

Do you want to update to fix a problem or just because it is the most current version?

If it is just because it is the current version, I recommend that you wait for the stable candidate version to become stable (17.15.3.1), as I noticed that the problem is not listed in that version.

[Ajesh]

We’re planning the firmware upgrade because one of the resolved issues in this version addresses a problem we’ve encountered:

“Resolved an issue that occasionally prevented the uplink port from being identified correctly on the Dashboard. This issue did not affect device connectivity.”

Since this fix is relevant to our environment, we believe proceeding with the upgrade is necessary to ensure accurate monitoring and management through the Dashboard.

[alemabrahao]

If it is something extremely necessary, go ahead. The release notes are information, so the bug may or may not occur. In this case, what I suggest is to update one location, monitor it for a few days and if everything is fine, update the other locations.

[Mloraditch]

I will also add you will often see a litany of bugs in Meraki and regular Cisco release notes. If a release is allowed out with an active bug it's generally going to be because that bug is rare. It's known and been identified via testing or support cases or else it wouldn't exist, but rare. You can always call support and ask for more details. They likely have internal notes about the exact cause or how many known reports exist that they can share in some fashion and then make a judgement based on your environment.

A non beta release is not going to have catastrophic or wildly debilitating known bugs.

[cmr]

What model(s) of C9300X are you planning to upgrade?  It might not apply to all.

[PhilipDAth]

C9300s are famous for this.  Nothing to do with Meraki.  It's an IOS-XE issue.

You should expect to reboot C9300s every 6 months to keep them running stably.  You'll probably be able to do this with the Meraki management system by doing regular firmware upgrades.

[GIdenJoe]

Which would be weird.  Virtually the same code runs on C9500's and I have a customer with a C9500 stack still running on 16.12 because they refuse to update due to 24/7 environment.  That stack has been running stable for years now.

[cmr]

@GIdenJoe I'm sure it'll very much depend on the exact elements you are using, I know on a particular Juniper firewall if you use the Anti malware feature it causes untold memory exhaustion issues, tha tmihht have recently been finally fixed...