v1.5.12 Bug Reports and Comments

[sakita]

Can confirm this behavior. In Netonix manager v1.0.20 multiple switches show a green ball and a very short uptime. Are they actually being polled? Is the status stuck? Don't know.

Three ways I have found to correct this:

- Disable polling on a switch. Wait a while and then re-enable it.
- Delete a switch from the list and then re-add it.
- Restart the Netonix manager (e.g. reboot the server)... and wait...

Also, when a switch is offline the manager will show a "Last seen:" date that makes no sense - e.g. a date from 4 months ago for a switch that was polling yesterday.

[jpmoller]

I suspect there is a memory leak somewhere.

While I am only running v1.5.11, given the notes on changes between this version and 1.5.12, I wouldn't expect there to be a difference between the two in terms of memory management.
I have three different models of switches running v1.5.11 and all three of them are exhibiting constant decrease in available memory, and when available memory gets to 0 and something needs memory allocated (like web UI login), the switch reboots itself. Not great when this happens at peak times during the day.

We see the exact same issue on switches running any firmwares older than v1.5.5 and has been fine up to v1.5.8. But seems to have come back since the string of firmware versions trying to patch TLS support. Graphs below to show available memory linear regression.

WS-8-150-DC (last 90 days) - board rev F:

WS-8-150-DC Memory Leak

WS-12-250-DC (last 90 days) - board rev F:

WS-12-250-DC Memory Leak

WS-26-400-IDC (last 90 days) - board rev B:

WS-26-400-IDC Memory Leak

The WS-8 and WS-26 are both new out of box.

[sirhc]

I am running v1.5.11 and v1.5.12 on many switches not seeing this at all???

If you have services loaded such as Discovery Tab or Discovery protocols and so on try disabling them 1 at a time and narrow it down.

[jpmoller]

Thanks for that.

We have discovery enabled on all of our switches on our network (177 in total), so would be odd that we are only experiencing it now on the new firmware (and the old firmwares mentioned).

But I will test disabling these one-by-one over the following week and get back to you with results.

[flameproof]

I am trying to upgrade a WS-26-500-DC, board rev B, PSU FW version 70, PSU board rev A, running on v1.5.8, and I see the following after uploading the bin file:

"This switch is using a newer temperature sensor because the older one has become obsolete, thus it is not compatible with the firmware that you attempted to upgrade too. Please only attempt to perform an upgrade with firmware 1.5.3 or greater".

The switch is running 1.5.8, which is greater than 1.5.3, is this expected?

[sirhc]

I am trying to upgrade a WS-26-500-DC, board rev B, PSU FW version 70, PSU board rev A, running on v1.5.8, and I see the following after uploading the bin file:

"This switch is using a newer temperature sensor because the older one has become obsolete, thus it is not compatible with the firmware that you attempted to upgrade too. Please only attempt to perform an upgrade with firmware 1.5.3 or greater".

The switch is running 1.5.8, which is greater than 1.5.3, is this expected?

There must be a logic fail in v1.5.8 upgrade routine, simply upgrade to v1.5.9 then upgrade from there.

v1.5.14 is also available now but you still have to first go to v1.5.9 as the upgrade process is of the current version installed on unit.