Netonix Forums

Is ERPS on Netonix G8032v2? Any help with matching up Calix options to Netonix Options would be helpful.

My Calix V2 options are:

Name:

Ring ID:
Interface-1:
Interface-2 :
Control VLAN:
Admin State:
Description:
MEG Level: Default is 1
Revertive Mode: ON/OFF Default is ON
Wait To Restore Time (min): Default is 5 min
Guard Time (ms): Default is 500ms
Monitor FOP-TO: ON/OFF Default is ON

Interface Config:
Interface-1: RPL none/owner/neighbor
Interface-2: RPL none/owner/neighbor
Netonix ?? Version Options

Instance - Not an option, auto generated. is this the same as Ring ID? You can't seem to set it which makes matching instances hard.
Level: Is this MEG level?
Control VLAN: Makes sense.
Monitor VLANs: No idea what this is or how to enter value(s)
WTR Time: Default 5 makes sense
Guard Time: Default is 500 makes sense
RPL: Bunch of options. Seems to be an odd way to set it but it works.






EAST Port: Makes sense
West Port: Makes sense

MEP: I have no idea what this is. Is it needed? A quick search shows it is used to signal up/down but I've never used it on a V2 Calix ring and it seems to monitor up/dow or a faded wireless link just fine.

Long term I would like to integrate with Calix, but for now, I can just do sub rings off my 10G Calix ring using little netonix rings if needed. Do I need MEP for anything?

Also, I am assuming you would want to set interfaces to duplex full, flow control off, STP disabled for any ring interfaces?

Much of the terminology we use can be defined here:
https://www.juniper.net/documentation/e ... ality.html

MEP's - Maintenance Association End Points, are point to point connections in the ring that monitor the links in the ring. If one of them goes down, then the RAPS (SF) is transmitted on the erps control vlan.

When building a ring, I usually start with the MEPs. Each switch in the ring will need two MEPS to monitor the east and west port for the adjacent switches.

petecarlson wrote:Instance - Not an option, auto generated. is this the same as Ring ID? You can't seem to set it which makes matching instances hard.

Good point, we can fix that.
Development on ERPS has been sparse because not enough people have been using it. Typically, our switches are on the edge of a network. Several major bug fixes/enhancements where done awhile back but where never confirmed beyond development testing.

petecarlson wrote:Level: Is this MEG level?

Correct.

petecarlson wrote:Also, I am assuming you would want to set interfaces to duplex full, flow control off, STP disabled for any ring interfaces?

Correct.

petecarilson wrote:I've never used it on a V2 Calix ring and it seems to monitor up/dow or a faded wireless link just fine.

Perhaps, but they do use them:
https://www.calix.com/content/calix/en/ ... ?42588.htm

Re MEG and Calix, you are right, but that is the wrong use case. For a Ring it would be:
https://www.calix.com/content/calix/en/ ... ?72460.htm

Is Netonix G8032v2

petecarlson wrote:Is Netonix G8032v2

I have spoken with the team about this to be certain.
Our implementation of ERPS is based on the G8032v2 standard.

I cannot say if the ring created by our switches is interoperable with Calix however because there has been no testing done for that scenario. If you would like to try and let us know how it goes here. It would be appreciated.

petecarlson wrote:Re MEG and Calix, you are right, but that is the wrong use case. For a Ring it would be:
https://www.calix.com/content/calix/en/ ... ?72460.htm

I'm afraid I don't understand what you mean.
Reading into the link you provided ( https://www.calix.com/content/calix/en/ ... ?72460.htm ) I found:

https://www.calix.com/content/calix/en/ ... ?72460.htm wrote:Fault monitoring

You may configure maintenance group end points (MEPs) on the individual ring ports for Continuity Check Message (CCM) monitoring (as defined in ITU Y.1731). MEPs represented on ring ports monitor each ring link. At a specified interval, MEPs send CCM protocol data unit (PDU) messages to check continuity on the ring.

...
Link failures

The RPL owner node may detect a signal failure condition and initiate a ring switch due to the following situations:

A fiber cut
Failure to receive consecutive timely or valid CCMs from MEPs
Receipt of a Signal Fail (SF) message from any node on the ring

This is precisely how we use them. I'm wondering if perhaps the Calix interface simply set's up the MEP's for you in the background. Whereas our's is designed to give you more control over it. I'm not familiar with Calix switches so I can only guess.