Unless you have TDD PTP links that fluctuate in available throughput wildly, sounds like a problem solvable by a grid of OSPF costs per PTP /30 matching different radio types. For example I have very different OSPF costs for each category of PTP link consisting of 80 GHz 1 Gbps FDD radios, Rocket M5s, Nanobeam M5-400s, 5.x GHz 256QAM TDD links (PTP600, etc).
On Thu, Oct 16, 2014 at 8:30 PM, Jason McKemie via Af <af@afmug.com> wrote: > This is pretty much what I've been thinking from the get go on this stuff. > The only advantage it has is being able to tell how much throughput is > available on the fly and assigning traffic accordingly. > > On Thursday, October 16, 2014, Eric Kuhnke via Af <af@afmug.com> wrote: > >> Solution in search of a problem. What did it claim to do that you can't >> do with a properly designed OSPF + BGP + MPLS (and VRF) capable network? >> >> Then again, I don't use Mikrotik routers, I use power hungry Cisco or >> Juniper gear so I don't have to deal with Mikrotik's wonky implementation >> of BGP and MPLS. >> >> On Thu, Oct 16, 2014 at 6:50 PM, Josh Reynolds via Af <af@afmug.com> >> wrote: >> >>> We pulled it off our network last week. Bunch of money down the drain. >>> >>> Not happy with the way Accedian handled this, on multiple fronts. It's >>> not like they were strapped for cash. >>> >>> Josh Reynolds, Chief Information Officer >>> SPITwSPOTS, www.spitwspots.com >>> On 10/16/2014 04:32 PM, Gino Villarini via Af wrote: >>> >>> We are going �to toss them >>> >>> Gino A. Villarini >>> @gvillarini >>> >>> >>> >>> On Oct 16, 2014, at 6:25 PM, Chris Wright via Af <af@afmug.com> wrote: >>> >>> 1.2.8.6_21016 released in June of this year. >>> http://forum.performantnetworks.com/threads/2015.1438/ >>> >>> � >>> >>> Chris Wright >>> >>> Velociter Wireless <http://www.velociter.net/> >>> >>> � >>> >>> *From:* Af [mailto:af-boun...@afmug.com] *On Behalf Of *Matthew Jenkins >>> via Af >>> *Sent:* Thursday, October 16, 2014 3:20 PM >>> *To:* Timothy D. McNabb via Af >>> *Subject:* Re: [AFMUG] Accedian/Performant/R-Flo woes >>> >>> � >>> >>> I have a tub of equipment I really wanted to deploy. It's not being >>> actively developed, so I have not deployed it. Are you running the latest >>> software? >>> >>> >>> >>> >>> -------- Original message -------- >>> From: "Timothy D. McNabb via Af" <af@afmug.com> >>> Date: 10/16/2014 1:58 PM (GMT-07:00) >>> To: af@afmug.com >>> Subject: Re: [AFMUG] Accedian/Performant/R-Flo woes >>> >>> Does anyone here on the list use Performant Nurons/Mind combination? >>> What did you find worked best? >>> >>> � >>> >>> -Tim >>> >>> � >>> >>> *From:* Af [mailto:af-boun...@afmug.com] *On Behalf Of *Chris Wright >>> via Af >>> *Sent:* Wednesday, October 15, 2014 5:00 PM >>> *To:* af@afmug.com >>> *Subject:* [AFMUG] Accedian/Performant/R-Flo woes >>> >>> � >>> >>> We�ve been sitting on a Performant Mind and four Nurons for almost a >>> year now. They sat for six months, then I spent another six months >>> tinkering with them here and there. Regardless of their EOL, we�re >>> looking to implement � but I�m running into the silliest of issues. I >>> can�t even get traffic from the Mind to a Nuron to pass without 20% >>> packet loss. This is without any routing, wireless backhauls, nothing. Just >>> a VLAN-tagged NIC plugged straight into the Mind, Ethernet from Mind to a >>> Nuron (where the vlan �pops�), then Ethernet to a second computer. >>> >>> � >>> >>> I�ve replaced the SFP adapters, the nuron, and verified none of the >>> Ethernet cables are bad by literally coupling them all down a line and >>> going straight from computer 1 to computer 2. No issue there. >>> >>> � >>> >>> This, combined with the fact that this stuff is going to be EOL in three >>> years, is maddening. My boss was under the impression that this should be >>> able to seamlessly integrate into our star topology (nuron at every tower >>> making a ring), but we can�t even get basic LAN functions to work >>> reliably. >>> >>> � >>> >>> Chris Wright >>> >>> Velociter Wireless <http://www.velociter.net/> >>> >>> � >>> >>> >>> >>
