Top Menu

Wednesday 18 Apr 2018 – Single upstream carrier Packet Loss Issue at NEW Data Centre

Wednesday 18 Apr 2018

Overview

At exactly 11.18am this morning, Mach monitoring systems detected issues concerning a single upstream carrier that was experiencing 5-15% packet-loss intermittently every minute or so.

All Mach infrastructure was first ruled out as having no issue nor any recent configuration changes.

The issue was confirmed to only be affecting a very small number of Mach clients, from only one of our Data Centres (Brisbane). For those clients, this most affected Remote Desktop (RDP) and other services acutely sensitive to packet loss.

Automated fail-over to Mach’s alternative link(s) did not reach thresholds of “link failure” due to the actual (modest and intermittent) level of issue. However, as Mach was preparing to intervene and make a manual forced change to direct this particular traffic flow through alternative upstream carriers – the root cause was resolved by them at 11.49am.

 

Further Details

  • Upstream Carrier notice:
    • Mach was advised late in the day that the error was caused by link congestion due to routing issues between carriers, which corresponds with what Mach experienced – i.e. intermittent packet loss due to an overloaded link segment
  • Thu 19th UPDATE:
    • On following morning, Mach also detected a similar but even lower packet loss issue between 8.39am and 8.59am
    • Potentially due to follow on engineering actions in response of yesterday
    • Mach will update this page, when further information available

Clients can see the latest up to date information of Status Notices online at: http://status.mach.com.au/

Comments are closed.