NZIX Engineers will be rebooting a switch to resolve a software bug stemming from the earlier incident: https://status.ix.nz/incidents/42rbl15t8mx4
Summary of impact during this window:
* Members connected to pe3.akl1 should expect their port to go offline.
* Members connected to pe1.akl1 should expect BGP sessions to drop to both rs1.akl.ix.nz and rs2.akl.ix.nz. BCP214 compliant filters will be applied on member ports here.
* ALL Members should expect connectivity to rs1.akl.ix.nz to go down.
Due to the nature of the software fault, we cannot appropriately install BCP214 compliant filters to ports on pe3.akl1 to reduce BGP disruption. This means ALL members that have routes via peers on pe3.akl1 may see a period of packetloss whilst BGP re-converges to alternate paths.
We have worked with the Vendor to try and resolve this without a reboot, but after investigation the CLI is not talking properly to the hardware and unfortunately we cannot try more invasive tests without risk of a watchdog initiated reboot. These riskier tests will be done at the time of the outage to provide debug to the vendor and provide us information moving forward should we expect this bug again.
NZIX has spare hardware on standby in the unlikely event this is a hardware issue, thus a 4 hour window has been set.
Expected Downtime: ~15 minutes
If Hardware Fault: up to 4 hours
Please contact email@example.com
or +64 9888 9221 if you have any queries regarding this notice.