I have noticed a high RPD usage.

show system processes extensive | match rpd

15755 root 102 0 12G 11G CPU1 1 702.1H 100.00% rpd{rpd}

15755 root 20 0 12G 11G kqread 4 73.7H 0.00% rpd{bgpio-0-th}

15755 root 20 0 12G 11G kqread 4 181:33 0.00% rpd{TraceThread}

15804 root 20 0 794M 24M kqread 4 2:20 0.00% rpdtmd

15755 root 20 0 12G 11G kqread 1 2:00 0.00% rpd{krtio-th}

Customer was observing intermittent RPD high, as high as 100%, on R1.

They were seeing interface flapping on R2, a router connected to R1, due to physical layer issues.

After cleaning the physical issue, the links have remained stable for R2.

And they were seeing the rpd issue in R1, exactly at the time when the above issue started & stopped once the above flapping was resolved.

When we see link flaps in connected router, there is every chance that route churns happen continuously.

R1 was getting routes from R2 via IGP. With this, the rpd will be go high to add/delete the routes in pfe/rpd resources.