[ixpmanager] Birdseye Looking Glass error version.

[Ikusan Charles]

Hi,

I followed your advice by using different vm for bird2 and ixpmanager but can't get the birdseye working.

I will appreciate if anyone could assist.

[Ikusan Charles](https://spikenow.com/r/a/?ref=spike-organic-signature&_ts=gs4xs)	[gs4xs]Warm Regards, Ikusan Charles A. Network Engineer / Monitoring | Internet eXchange Point of Nigeria IXPN Office: 8th Floor, NCR Building, 6 Broad Street, Marina, Lagos. Phone:+234 (0)700 CALL-IXPN, +234 (0)8035606131, +234 (0)8093878116 [www.ixp.net.ng](http://www.ixp.net.ng/) | noc at ixp.net.ng | www.twitter.com/IXPN_NG | www.facebook.com/IXPNigeria

On April 14, 2020 at 14:22 GMT, Ikusan Charles <c.ikusan at ixp.net.ng> wrote:

Thanks Nick, I know the source of my issues now

[Ikusan Charles](https://spikenow.com/r/a/?ref=spike-organic-signature&_ts=griym)	[griym]Warm Regards, Ikusan Charles A. Network Engineer / Monitoring | Internet eXchange Point of Nigeria IXPN Office: 8th Floor, NCR Building, 6 Broad Street, Marina, Lagos. Phone:+234 (0)700 CALL-IXPN, +234 (0)8035606131, +234 (0)8093878116 [www.ixp.net.ng](http://www.ixp.net.ng/) | noc at ixp.net.ng | www.twitter.com/IXPN_NG | www.facebook.com/IXPNigeria

On April 14, 2020 at 14:15 GMT, Nick Hilliard INEX <nick at inex.ie> wrote:

Ikusan Charles Abimbola wrote on 14/04/2020 13:31:
> Lastly, my concern is that, I hope lighttpd won't affect my apache2
> because I installed my bird2, birdseye and ixpmanager into one Linux box.

lighttpd will fight with apache2. This won't ever work properly.

You need to install route servers on their own separate virtual machines.

You also need two route servers. The reason you need two is that
sometimes you will need to upgrade one, so you need the other one
running while you do this.

People often ask us what we use at INEX, so here's roughly our layout:

1x physical server / hypervisor running primary route servers in a VM
1x physical server / hypervisor running secondary route servers in a VM

1x hypervisor cluster running everything else:
- ixpmanager vm
- database vm
- sflow vm
- graphing vm
- other stuff

The reason we keep the route servers on their own physical servers is
because route servers are considered production IXP services, i.e. they
relate causing packets to flow across the IXP peering platform.

Everything else is management, so that's on its own separate hypervisor
cluster.

Nick
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.inex.ie/pipermail/ixpmanager/attachments/20200414/1f35e590/attachment.htm>