[9.070][BUG] Memory access error

Hi,
I have been trying to determine why my external link randomly drops the ethernet connection for 3 seconds. For some strange reason a number of other UTM functions fail. The break is too short for the PPPoE service to see and restart.
The external interface is an Intel onboard NIC.

All I can find is this


2013:02:26-23:09:33 mobile-cats kernel: [404186.674162] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready 
2013:02:26-23:09:47 mobile-cats kernel: [404200.779284] Program lshw tried to access /dev/mem between ff000->101000.

Ian
The users are complaining of slow response which I think is explained by the link failing.

This interface is using the IPv6 native stack and PPPoE.

Parents

0 Ol Deda over 12 years ago

https://community.sophos.com/products/unified-threat-management/astaroorg/f/80/t/64971

https://community.sophos.com/products/unified-threat-management/astaroorg/f/80/t/64966
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 Tinkerbell over 12 years ago in reply to Ol Deda

Hi,

This post is not a duplicate. The threads mentioned currently deal with the following:
https://community.sophos.com/products/unified-threat-management/astaroorg/f/80/t/64971
The bug here deals with webadmin being not responsive when opening live log.

https://community.sophos.com/products/unified-threat-management/astaroorg/f/80/t/64966
This is DNS issue with PPPoE and IPv4

This is another scenario, using Ipv6 over PPPoE. Will talk to dev team and check what they say regarding these logs.
Regards!
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel

Reply

0 Tinkerbell over 12 years ago in reply to Ol Deda

Hi,

This post is not a duplicate. The threads mentioned currently deal with the following:
https://community.sophos.com/products/unified-threat-management/astaroorg/f/80/t/64971
The bug here deals with webadmin being not responsive when opening live log.

https://community.sophos.com/products/unified-threat-management/astaroorg/f/80/t/64966
This is DNS issue with PPPoE and IPv4

This is another scenario, using Ipv6 over PPPoE. Will talk to dev team and check what they say regarding these logs.
Regards!
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel

Children

0 RFCat_vk_01 over 12 years ago in reply to Tinkerbell

Hi Bianca,
so far today the log shows nothing, no entries, but the condition didn't start until late yesterday.
In the meantime, I have changed the ethernet cable to the modem and restarted the modem. The weather here has been warm (around the low 30c mark) with very high humidty for the last couple of days. I have shifted the modem to allow better airflow to it, mind you it is a new box with modem only.
I will keep an eye on the logs etc during the day.

Ian
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 RFCat_vk_01 over 12 years ago in reply to RFCat_vk_01

Nearly 24 hours and no new log entries.

I will wait until tomorrow to declare it a non event. Not a myth though.

Ian
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 RFCat_vk_01 over 12 years ago in reply to RFCat_vk_01

Hi Bianca,
So far tonight no new log entries. I have thrown the ethernet cable out, not going to experiment with it.
The condition did cause the UTM to stop functioning, a number of functions failed and restarted and the internal interface would not accept new connections.

Anyway, it isn't a myth, but not sure what to call it?

Ian[[[:)]]][[[:)]]][[[:)]]]
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 Tinkerbell over 12 years ago in reply to RFCat_vk_01

Hi Ian,

I assume it was a hardware malfunction.
Usually we call a myth thread in which the scenario which is not reproducible. So i will set the title like that for now. If you get any new log entries regarding this please post them.
Best regards,
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 BarryG over 12 years ago in reply to Tinkerbell

Usually we call a myth thread in which the scenario which is not reproducible.

(OT)

I really don't like the use of the word 'myth' here. I don't know if this holds true in German for 'mythos', but in English, 'myth' can mean 'a widely held but false belief/idea'
Definition of myth in Oxford Dictionaries (British & World English)

ISTM that 'unreproducible' or similar would be a better tag to use.
For things that are confirmed to be user error or by design, you can use 'NOTABUG' or similar.

Barry
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 RFCat_vk_01 over 12 years ago in reply to BarryG

I have to agree with Barry, in this case definitely is not a myth.

While it might be a hardware failure external to the UTM, it is a bug that brings the UTM to its knees each time the ethernet cable fails, the UTM does not accept connections, middleware restarts so do a number of other functions. The break is not long enough for the ADSL connection to drop.

The UTM should be able to function if one interface fails or has an intermittent problems.

Ian
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 Tinkerbell over 12 years ago in reply to RFCat_vk_01

Sorry guys if the word Myth is disturbing. It is a tag used in BetaBot for the situations I mentioned . I will try and rectify it thought in the future [:)].
Ian i saw this error again in kernel.log you attached in the post : https://community.sophos.com/products/unified-threat-management/astaroorg/f/80/t/64985

This error in lshw apparently appears in the moment you are breaking the link.
So far all related PPPoE issues/threads are tracked in the mantis ID associated with the link above.
Keep you posted. Cheers,
Bianca
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel