Hi all, after a power failure my stand alone 3560 shows a config.text.renamed and a vlan.dat.renamed instead it's original files.
Thanks a lot
I've seen this before and it's usually a result of a feature called express setup.
If the mode button was pressed for >10 seconds then what the switch will do is rename the config it as it has done here.
You can see the document about this for a 2970 however it's the same for a 3560:
You can disable this feature by typing "no setup express" in the global configuration. A lot of our customers choose to do this to prevent someone walking up to the switch and holding in the mode button.
Hope this helps,
Cisco TAC - Campus LAN
Many thanks Michael for your prompt answer.
I'll certainly try it.
....but can it happens without pressing mode button.
Thank you a lot again
It can't really happen without pressing the mode button, as the developers have looked closely at the code to see if there is a problem. I guess if the button is stuck down then this could cause it too.
Has the problem occured since?
Our 3560 is POE capable and we have some 7960 phones plugged on it.
It seems to loose power and reloads... and after reload we find those two files renamed
We tried to change the device but the behaviour is the same. The power of our rack is stable.....
Many thanks again
The bug i'm looking at is CSCee75743 which is marked unreproducible (i.e. all the cases reported like yours the developers have not been able to reproduce). In fact, Cisco had reclaimed a few of these switches from previous cases and did extensive tests on them to see if environmental factors like heat or cold could trigger the button to be set.
How often do the reloads happen, and do they occur at any particular time of day? Is the rack the switch is in locked?
I would therefore firstly look to see if anyone is pressing the mode button and hence setting this problem off.
I don't think it would be power related. However, I did a test the other day on a 3560 and pressed the button of the switch down for 12 seconds. The switch did reload and the files were renamed like you saw too.
What I would suggest is to configure the command "no setup express". You should find that the problem will not reoccur once this is in place.
Do you have a way of reproducing this problem that Cisco can try? This would help greatly as none of us have been able to recreate the problem here.
This issue happens randomly two or three times a day. Our customer is a Bank and the access to the rack is allowed to authorized personnel only. I've been there for 9 hours with my pc connected via console...
The only thing I've seen is a "reload by power on" message. I'll be there angain tomorrow morning and I'll check for power absorbed by phones, maybe one or more phones are experiencing a power issue.
I'll let you know.
I'm seeing the same behaviour with a
WS-C3560G-24TS running on 12.2(25)SEE2 (C3560-IPBASEK9-M). No PoE, but still the same issue. I can also rule out that somebody pressed the "Mode" button.
Did anybody find the root cause to this problem so far?
There is no root cause Cisco can find.
Are you able to reproduce it? Was the power to the device cut, or did it reload on its own accord?
The workaround is to configure "no setup express".
If this is reproducible please let me know.
I am having the same issue. We are deploying about 200 2960-24 port switches in which about 75% of those deployed so far have been randomly rebooting and going into express setup. I have noticed the filenames for the configs were renamed as mentioned by previous messages posted. When I rename the file to config.text and reload all is restored.The same thing has also happened on a 2950-24 port switch at one of our sites, which I found odd.
I am wondering if this is a bug in the 12.2(25)SEE2 code since it does not seem to be specific to a certain switch model.
It also seems to happen when we are imaging checkout lanes through the network from a server on the same LAN. Could excessive input errors cause a switch to lose it's config? In some cases there was a speed/duplex mismatch with the server being set to auto and the switch set to 100 full. Also the server is running VMWARE with virtual switching enabled on the NICs.