OSPF bad checksum problem

[jonathan.chapman]

I would welcome comments on the following scenario.

I am seeing OSPF bad checksum error messages several times an hour between two OSPF adjacent routers. Both routers have other adjacencies established with no problems. Looking at the physical interface there are no errors and the FR has no problems , indeed both routers have other non-erroring adjacencies via the same physical links.

One router is running 11.2.14 and the other 12.2.6

[steve.barlow]

Tough cause to find, use this link to help you narrow it down: http://www.cisco.com/en/US/tech/tk648/tk365/technologies_tech_note09186a0080094819.shtml#2

Steps include, check cable, change switch port, directly connect the routers (to eliminate network as a cause), then call TAC.

Hope it helps.

Steve

[jonathan.chapman]

Hi Steve,

I had already seen this doco but thanks for your comments anyway. I think we can eliminate the physical side because I have other adjancencies running ok via the same cable/ csu/ serial conx with no problems.

But if you have any other ideas.....

[thisisshanky]

What IOS are you running, on the routers on which adjacancies are working fine ???

The link that Steve provided, does say that, the FR switch could corrupt the packets....Since you are running frame-relay you will be having a few pvcs terminating at your router. You might want to check if you see any errors on this particular DLCI, ...Even the LEC can be asked to run test patterns on this DLCI to see whether data transmission over this PVC is working fine, without errors.

ALso it could be due to a bug in the IOS...ultimately you might want to open a case with the TAC.

[jonathan.chapman]

The IOS we are running on the adjacencies that are ok would be 11.2.18

I think that if there was a problem on the FR side this would be reflected in the PVC stats on the router, also I would see errors on the serial interface stats. But I am seeing the serial interface and the PVC stats as being clean. Unless I am mistaken.......

I could open a ticket with the TAC but I doubt they would know and I would end up with the standard Cisco "we dont know, upgrade your IOS" response.

[thisisshanky]

I would go for opening a TAC case, as you have got nothing to lose. It could really be a problem due to an IOS bug, as all the 3 router IOS you mentioned were different.

[jonathan.chapman]

Shanky et al thanks for taking the time to reply to my posts. This problem is now resolved. Here is the final troubleshooting solution for anyone that is interested - Although all the physical links were showing clean I asked the PTT to rebuild the PVC using a different switch port on the IGX and this has resolved the problem ,although for the moment I cannot be sure what the cause of the problem was I think it could have been a case that the PTT upgraded switch card and the new card has a bug which created the problem (although this is just a hunch).