[AusNOG] Carrier aggregation port redundancy

[Mark Tees]

True. The motivation is probably somewhat limited by the options that are
already there in terms of more expensive services. Why would $T or $A
change that unless they were absolutely forced by the market to do so?

On Tuesday, August 18, 2015, paul+ausnog at oxygennetworks.com.au <
paul+ausnog at oxygennetworks.com.au> wrote:

> I agree Mark that L2TP isn’t an issue, and DSL or something is an obvious
> choice for a backup of a customer EVC, but I am more concerned about
> aggregation port redundancy, protection for our business not just some
> customers.
>
>
>
> We already have multiple aggregation ports due to capacity requirements
> and the carriers preference not to go 10G, go figure, so it’s already
> getting complicated from that point, but really, apart from covering your
> own equipment failure, having redundant ports in the same DC Is a waste of
> money and not really an adequate DR plan for our business as far as I’m
> concerned.
>
>
>
> We can easily split our core between geographic locations, we can take
> peering, transit, and L2TP handoffs for DSL and NBN stuff in multiple
> locations easily, but redundant geographically separated EVC/VLAN handoff
> for Ethernet services seems to be an unnecessary challenge for carriers,
> when really it should be designed into a product from the start.
>
>
>
> Ordering second EVC’s would be an answer for sure, but ultimately you are
> doubling your cost everywhere unless you are happy for a lower speed EVC
> for redundancy, and with the response times and restoration times seeming
> to be impossible to meet for carriers lately we can’t afford to run some
> remote sites on low speed links for 12 hours while Telstra gets out of bed
> at 2am.
>
>
>
> Regards
>
> Paul
>
>
>
> *From:* Mark Tees [mailto:marktees at gmail.com
> <javascript:_e(%7B%7D,'cvml','marktees at gmail.com');>]
> *Sent:* Tuesday, 18 August 2015 12:10 AM
> *To:* Nathan Brookfield
> *Cc:* paul+ausnog at oxygennetworks.com.au
> <javascript:_e(%7B%7D,'cvml','paul%2Bausnog at oxygennetworks.com.au');>;
> ausnog at lists.ausnog.net
> <javascript:_e(%7B%7D,'cvml','ausnog at lists.ausnog.net');>
> *Subject:* Re: [AusNOG] Carrier aggregation port redundancy
>
>
>
> L2TP terminated stuff should be easy enough providing the carrier lets you
> advertise your LNS prefixes by BGP etc.
>
>
>
> Layer 2 hand offs could be redundant psudowires or VPLS multi-homing in
> theory but who is going to do that when they can bill for another EVC or
> specifically a VPLS service :D
>
>
>
> For sites that require that level of redundancy it's probably good to
> build in a backup via DSL or a second EVC.
>
>
>
> Also, having DR plans for those links in place is good too. IE have a
> second port on another device prepped and get remote hands to move it if
> there is a major failure on your side.
>
>
>
>
> On Monday, August 17, 2015, Nathan Brookfield <
> Nathan.Brookfield at simtronic.com.au
> <javascript:_e(%7B%7D,'cvml','Nathan.Brookfield at simtronic.com.au');>>
> wrote:
>
> Hey Mate,
>
>
>
> Layer 3 redundancy everyone will do without to much pushing, Layer 2 no
> one will.
>
>
>
> If customers want Layer 2 redundancy they need EVC's to multiple DC's.
>
>
> Nathan Brookfield
>
> Chief Executive Officer
>
>
>
> Simtronic Technologies Pty Ltd
>
> http://www.simtronic.com.au
>
>
> On 17 Aug 2015, at 23:11, "paul+ausnog at oxygennetworks.com.au" <
> paul+ausnog at oxygennetworks.com.au> wrote:
>
> HI all, just wondering what experience people have had with negotiating
> aggregation port redundancy/failover at different datacentres within the
> same state for major carriers ?
>
>
>
> We are looking at implementing some geographic redundancy for our layer 2
> connections which are handed off on an aggregation port in one datacentre
> so that we have a second aggregation port in another datacentre should the
> carrier or ourselves experience a major failure like an exchange switch or
> something.
>
>
>
> We have had 2 or 3 failures over the last couple of years which luckily
> were on weekends and the impact was lower than normal, however the carrier
> does not have any interest in providing such an option across datacentres,
> they will do it within the one datacentre as long as you pay the full
> install and MRC on the second redundant port, but I can’t get them to do
> anything in separate locations.
>
>
>
> Am I the only one who thinks that having geographic redundancy for such
> services is important ?
>
> They don’t seem to think it’s an issue and are saying that they won’t do
> it but I would really like to have it for obvious reasons and am not sure
> if I should keep pushing or not.
>
>
>
> I’m interested in other peoples experience in this area please.
>
>
>
> Regards
>
> Paul
>
> _______________________________________________
> AusNOG mailing list
> AusNOG at lists.ausnog.net
> http://lists.ausnog.net/mailman/listinfo/ausnog
>
>
>
> --
> Regards,
>
> Mark L. Tees
>


-- 
Regards,

Mark L. Tees
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ausnog.net/pipermail/ausnog/attachments/20150818/9634b81c/attachment.html>