Hello,
This question relates to what might cause several IP SLA operations to affect each other when set up on the same device.
Our setup involves several model C6500's and up. The links between these devices are a mix of technologies. To ensure that the various links with their underlying leased lines etc. are up and running (the leased line may be down, but the interfaces are all registering as UP on the Cisco devices) we use IP SLA UDP jitter operations.
Last night we had one leased line go down, which quite correctly registered on the correct IP SLA operation as was expected. However, this specific devices also had another IP SLA operation configured, which monitors a different link, which is even running over a completely different technology (so no common denominator outside the cisco device).
The operation I expected to see the outage register on, gave this error message:
Operation did not occur because no connection (session) exists with the target.
While the one that monitors the other, unaffected link, gave this error message:
Socket failures or other errors not relevant to the actual probe.
First of all, what exactly do these messages mean?
Secondly, the only common thing I can find is the port number. Source port is 5001, and destination port is 5000, for both operations, but the IPs are different of course. Maybe the socket error relates to the identical port numbers, and that there is an issue with port/IP mapping? I have tried to find out what these error message may translate to in "plain english", but have found nothing that gets me closer to an explanation for why the outage registered with the same identical time frame for both operations.
Hope someone can point me in the right direction here..
- Ole
