ISR4451 error every 6 seconds: dash_i2c_master_xfer: I2C Master Error

[portmannth]

Hello,

One of our ISR4451 started, some time ago, to log the following error every 6 seconds (its uptime at this time was 3 years and 37 weeks, running 16.6.7):

Aug 27 03:29:41 MEST: %IOSXE-3-PLATFORM:kernel: dash_i2c_master_xfer: I2C Master Error master=0 addr=0x75 status=0x15

Trying a reload resulted in a non functional router. The PS LEDs where green, but on the front, nothing was lit. Doing a cold power (including taking it off from the power outlet allowed the router to come back to life and boot like nothing ever was wrong.

A search for the logged error just gives me nothing useful. Does anyone here know more about it? Is this message due to some kernel problem and the cold boot fixed it, or is it rather an indication of a hardware problem and the router should be replaced?

Thank you in advance and best regards

Tom

[Giuseppe Larosa]

Hello @portmannth ,

>> its uptime at this time was 3 years and 37 weeks, running 16.6.7):

>>

Aug 27 03:29:41 MEST: %IOSXE-3-PLATFORM:kernel: dash_i2c_master_xfer: I2C Master Error master=0 addr=0x75 status=0x15

The error is related to the kernel and you solved with a power cycle of the router if my understanding is correct it should not be a sign of an hardware problem.

Some SW issues arise only after very long uptime like yours.

Keep an eye on the router and see how it behaves in the next days

Hope to help

Giuseppe

[Leo Laohoo]

---

@portmannth wrote:
its uptime at this time was 3 years and 37 weeks, running 16.6.7

---

The uptime is one reason.  Never leave a router, particularly one that is IOS-XE, running for more than 1 year without a reboot.  

However, I think this is the cause of the memory leak:  FN - 72323 - Cisco IOS XE Software: QuoVadis Root CA 2 Decommission Might Affect Smart Licensing, Smart Call Home, and Other Functionality.  In 6 months time, check the memory utilization for the "keyman" process using the command "sh process memory platform sort location rp".  The "keyman" process should have a memory utilization of <15000.  Anything higher than that means trouble.

Look at the logs.  If the logs is constantly being spammed with Smart License failures then manually upgrade the certificate as the least painful workaround.  

[portmannth]

Thank you both. I will leave this router running as is and see if the problem arises again this next few days.