Reports of LexisNexis service problems included DNS failures for both service endpoints and the status page. How should a status page be set up so customers can still get outage information?
Why put a status page on the same domain as the service?
Asked by
!weigh~the+paperweight
Answers
A separate hostname is not necessarily enough. If the status page and service depend on the same authoritative DNS, a DNS failure can take both names out of reach even when the status page itself is still running.
Use a separate domain with independent DNS, and check that its hosting and other critical dependencies are separate too. Then test it by simulating a failure of the primary domain or infrastructure.
Reply to this answer
There were differing reports about what was reachable. One account described widespread DNS failures, while another reported being able to log in to a service. That does not establish that every product or location was down.
A report also described reaching the status page through a direct hostname override, where it showed a partial outage. That suggests the page could be running while normal name resolution was failing, though it does not establish the root cause.
Reply to this answer
Have a backup communication channel as well. Email alerts or another independently reachable notice can tell customers what is affected and when to expect another update if the status page itself cannot be reached.
Reply to this answer
Add your answer
Answers are reviewed before publication.
Kedric.8w:
Also make sure customers know the separate address before an outage. A status page is less useful if the only link to it is on the unavailable site.