May 2, 2017 · Hosting, News · Comments Off on Hawk Server outage

Unfortunately due to a hosting platform failure on Sunday morning (30 April) at the data centre the Hawk server is currently offline.
This outage affects my primary website and all email communication as well as client websites and email that are hosted on the server.
I will post updates here as they become available.

Latest Update:

2017-05-05 3:00pm
All database issues have been corrected and affected tables have been restored.
No further issues have been identified.

2017-05-03 09:00am
A few sites have reported database connection issues.
We are working to correct the issue and restore the affected database records from backup.

2017-05-02 1:55pm

All sites seems to be responding normally and email is coming through.
Emails sent while the server was down should have been held by the sending server and will now be accepted and sent through.
They would have tried for 24 hours before bouncing back to the sender (some may have a shorter time).

Server load is up a bit at the moment as expected but is settling down as traffic drops back to normal.

2017-05-02 1:40pm
The server is back online and serving traffic and all sites restored from latest backup.
I am currently running through all sites to check the status.

2017-05-02 08:30am
The data centre are proceeding with the Bare Metal Restore of the server.
No Estimated Time of Repair is available at this time.

Outline of issue:

At approximately 10 am on Sunday morning the Hawk server stopped responding, upon investigation it was found the server administration panel could also not be accessed so the problem was reported to the data centre to follow up as it was most likely a hardware issue and not just the VPS (Virtual Private Server) that was offline.

The data centre confirmed this and proceeded to investigate.

At approximately 12pm the Systems Engineers provided addition feedback that the issue was more widespread that first through and affected the entire Xen platform itself.
Work continued to identify the root cause of the problem and find a solution.

By Monday morning they had identified the problem and determined that the existing platform could not be recovered to a working state.
The only option was to perform a Bare Metal Restore for all affected services.

What is a Bare Metal Restore:

Effectively the data centre need to set up all hardware from scratch as a new server framework.
They will then load all required software and test the set up.
They can then configure the new framework for the Virtual servers and then restore the VPS including all accounts from backups where available.

For my servers I have managed backups in-place that keep a rolling 14 day backup of the server just for this purpose.

How long will this take:

Unknown at this time as I am unsure of how much hardware needs to be set up or the scope of the data that needs to be restored.
(information of this nature is commercial in confidence.)
The best people for the job are on it and working as quickly as they can to restore services and will let me know when my server is being restored and available so I can test everything is back the way it should be.

Thank You
Ron Rogers
WEBY Systems

extra line goes here