Log in to ask questions, share your expertise, or stay connected to content you value. Don’t have a login? Learn how to become a member.
I am running vMX in a VMware vCloud environment and I am running into a very high CPU issue with eventd running at over 45%. In addition, I am seeing the following log messages over 1000 times a second. I disabled the SYSLOG file messages but the high CPU remains. I have restarted fpc0 to no avail. In addition, I am running the same version 19.3.2 in VMware vSphere environment and do not see the issue. It appears to be unique to vCloud but there is not much I can tweak in vCloud. I have attached the support logs if anyone is up for taking a look and I have opened a JTAC case. However this is a critical showcase lab for JUNOS and SR-MPLS and looking to get this resolved. One last note when i was messing with clearing out the config I lost my fpc connection between the VCP and VFP and at that point the high CPU stopped. I was forced to restart the VMs to restore the connectivity and the issue returned when the VCP-VFP connectivity returned.
The said messages are due to the interface type selected.
Confirm if the interface type is VMXNET3 and not E1000
Thanks so much that worked. I was working this case with JTAC escalation and they were not aware of the root cause of this error. Befero contacting JTAC, and in order to get more than one GE data interface working in vCloud you need to delete all the interfaces and re-add them and not enable NIC 0. If you do not do so the MAC address assignment gets shifted and the INT interface in the VFP does not correspond to the MAC in vCloud on NIC 1 (2nd NIC) and the FPC0 does not come up. It took a lot of testing on vApp to figure that out, however when you re-add them the default interface type is E1000 which is what caused the problem.
Can you paste the current config?