Ever seen this: %% Low on memory; try again later

[Lsimancek_2]

I have a 3750 switch that currently will only display this error:

%% Low on memory; try again later

The only warning/error that is on our logging server previous to this is a traceback error. Thier is no history of this on this stack, and if it was so low on memory I would have expected to see a %memalloc error before a full failure.

Any ideas? At the very least I'm trying to figure out if I have more going on here then simply memory overrun and needing a reboot.

[Giuseppe Larosa · Accepted solution]

Hello Lsimancek,

this message should mean that unfortunately it is high time to reload this device.

I've seen this message on routers. On a switch like yours it should be the sign of a SW bug.

Hope to help

Giuseppe

[Lsimancek_2]

Giuseppe,

Ouch, figured ... reboot it is!

Now here is part 2: This switch is apart of a 2 switch stack, the switch with this error is the master switch, what is the likelyhood that the master switch is the only one running out of memory and the member switch will take over the stack when the current master dies or is rebooted without problem?

[Leo Laohoo]

I've seen this error message on a WS-X4013+ and all traffic just ground to a halt.  According to TAC it was a bug in the IOS and I had to upgrade the firmware to fix the issue.

[Lsimancek_2]

I haven't found a doc yet nor does anyone around here know, so do you guys know if the master switch is having the memory failure should the member switch be fine and take over? Or does having the master switch "lose its mind" negativily effect the member switch?

[glen.grant · Accepted solution]

    The other switch will be fine but it will become the stackmaster and it will not fall back to the original switch by itself .   If that is a concern make sure the switch priorities are set  and just reload the whole stack .   We have seen the low memory issue also on a 6 switch stack , we reloaded the whole stack . I think the stackmaster sometimes runs into issues  because it  has to handle  the configuration and tracking of things like spanning tree and other functions for the whole stack .  My take it is a bug and for some reason these  eventually run into a memory leak situation.   Cisco should look into this further but I can't say I have seen a bug  for this and I have looked before .  We run like 12.2.35SE  .

[Lsimancek_2]

Unfortunately since I can't get in I have no idea if the priorities were set correctly. These were installed by a co-worker long gone. I hope they are, right now our plan is to let them run until wither it reloads itself or I can this weekend.

Thanks for the help!

[Leo Laohoo]

The command to check the priorities is "sh switch detail".

If you need to change the priorities it's "switch priority <1 to 15 with 1 as default>".

[Lsimancek_2]

Yes, but right now the switch is pretty much useless (other then its still passing traffic)  because from console you can't get a prompt. All you see is

     %% Low on memory; try again later 

when attempting to use the switch. The console on the other switch in the stack is unresponsive, so I can't access the device to alter priorities. But thanks for trying

[Luis Antonio Fernandez]

Hello.

I have the same problem on a C2960-LANBASEK9-M Version 15.0(2)SE11 switch and it has been a long time since 2009.

[bbb bbb]

Dears,

Encountered the same issue as of this date. A reboot will usually fix the issue. IOS upgrade is not feasible as it is end-of-support.

It is good that network traffic is not impacted and now waiting for the weekend (non-working days) for a quick restart .:D

Best regards