Internet Exchanges: Peering go-live over 2x IXPs

Updated

We’ve now deployed a dedicated Emergency Access Circuit into our Core Network.

This is an optical fibre connection purely for OOB (Out-of-Band) access in emergencies.
It’s low-bandwidth yet 24/7, giving us a redundant path in that doesn’t need Cellular etc.

Ended
Resolved

We’ve successfully completed the Core Router reboots. Thank you for your patience.

Please let us know if you’re seeing any routing issues/delays so we can investigate.

Further minor internal adjustments will continue to be made post-advisory.

Monitoring

Internal BGP over IPv6 has been restored, after Router firmware upgrades changed default behaviour.

This means our Core Routers are now once again able to exchange routes over IPv6 as well as over IPv4.

Ultimately the key benefit of that is that a route which exists over 1x IX can be taken by either Router.

We’re keeping this incident open as we need to restart the Routers to enact affinity config updates.

Monitoring

Both of the Internet Exchange Points are now live in our Core Network! We observed no outages.

As we’re using Route Servers for multilateral peering (MLPE) no further work is required on our end.

We’re remaining on-site in case of issues, and are now working on network house-keeping works.

Note: If you’re not seeing improved routing, please ticket us with ISP and IPv4/v6 info, etc. :-)

Updated

We’re cautiously beginning routing/filtering amendment works, with engineers on-site.

Meanwhile, the emergency circuit order has been progressed by the facility.

Updated

All 8x eBGP sessions are now established. Routing will be amended tomorrow when we’re back on-site.

Further to that, we’ve lodged a request to deploy a dedicated out-of-band emergency circuit into our core.

Began
Maintenance Planned

Hi there,

After much planning & ground-work, we’re going live with 2x Internet Exchange Points (peering fabrics)!

We’ve established contact with the 2x Route Servers per-IXP already. This advisory is bringing up sessions, etc.

  • Once we’ve got the sessions live, we’ll then configure path/route costing to prefer peering over transit
  • Due to using Route Servers rather than explicit bilateral peering per-ISP, this will quickly improve routing
  • There’s a small likelihood of some routing blips during these works (while we iron out new routing profiles)

Please let us know if you’ve any queries or concerns. Thank you for your patience as we improve routing.

Cheers,
Merlot Digital

10 Affected Services:
The Network Crew Pty Ltd (TNC)

« Merlot Digital's website «

Network: AS138521