Using Peplink For core routing in a Data Center

Hello ,

we are a cloud service provider and We are thinking of adding the balance 2500 as our core router,
the main reason is because the outbound policy manipulation peplink is able to do is unbeatable,
we are using now Mikrotik, we run about 300MB , and we run multiple BGP peers,
We couldn’t find any major case studies using it as a core router,
we use a tone of the lower peplink balance models for our customers, but we want to know if the 2500 is even more stable than the balance 20,

Confidence from users who actually deployed it, will help much

thank you,

We are a cloud service provider and have been using peplink as our core/border router in several data centers for five years. We provide ip phone service and cellular backup Internet to pizza restaurants using MAX-BR1 or equiv at customer location, each with two speedfusion/pepvpn connections to two data centers for redundancy.
We are currently running 850 vpn connections from each data center.
The entire upper balance series is very stable. From the 710 to the 2500 we have not had any issues at all for last two years (well…I did test a beta firmware and found a bug, but nothing in release firmware).
The high availability mode works very well. You can have a failover pair of 2500s. config changes to master flow to slave automatically. If master hangs slave takes over in a few seconds.

3 Likes

thanks mush for your detailed response, this is really encouraging!,
do you use BGP ?
do you use teh SD switches ?
how about emergency support from peplink,

We do not use BGP, as we use a high quality five carrier blend from the data center.
We do not use the SD switches. our internal switching fabric is all 10G. We use a high availability virtualization platform called Proxmox, and we have six 10G switches connecting the physical servers. two for primary and failover CEPH (distributed file system for VM images), two for primary and failover cluster (inter server) communication), two for primary and failover “public and private IPs” for the VM’s IPs.
Emergency support is pretty good. We never need to call them though…with almost 900 devices I think we have called support once in the last 90 days. I had not thought about it, but I think the 8.x firmware has been an order of magnitude improvement in stability - not that it was bad before. Just really stable now.
Also - Peplink has been very responsive to enhancement requests. I have made quite a few, and a good percentage have been implemented.

6 Likes

thanks mush for your responds !,