Hey guys,
Neither Level3 or Juniper could find a smoking gun, so I'm curious if anyone's ever ran into this behavior.
I have 3 sites with pairs of MX104s (running 17.4R2) in each site and waves between the sites (ring topology). Periodically the link between a given pair of sites goes down for no good reason, despite having good light levels. In order to bring the link back up I have to offline/online the entire card on one or sometimes both sides, since simply disabling/enabling the interface doesn't seem to do the trick. The link down event does also occur for good reasons occasionally like a fiber cut and link state doesn't always recover without intervention.
In this situation we're working with 1 Gb waves delivered with SMF, and configuration matches between all sites. Once the links are up traffic passes through just fine, but whatever blip causes the ports to go down every few weeks results in the ports not recovering without intervention. Nothing strange pops up in the logs. The location that most frequently experiences this issue does have the colo provider running DWDM between the buildings where Level3's handoff is and where our equipment is.
There's also a non-zero chance I should've just escalated all the things until either vendor could find something wrong, but their conclusions line up with everything I was seeing.
Has anyone ever ran into something like this before?
TL;DR: 1 GbE SMF ports on MX104s connected with a Level3 wave remain in a down state despite having good light levels on both sides, until the card gets offlined/onlined. Support from both vendors doesn't see anything wrong.
No comments:
Post a Comment