Quote:
Originally Posted by General Maximus
Not exactly, there are various tests you can perform to see if a problem might be a dns issue, such as pinging the dns server or submitting a query
|
Actually some moderately comprehensive DNS tests are included as part of
Netalyzr. These test both forward and reverse lookups of a collection of popular domains, response times, reliability, whether your DNS server can correctly reach all the root servers, and also how your DNS server responds to odd and malformed requests.
---------- Post added at 14:10 ---------- Previous post was at 13:02 ----------
Quote:
Originally Posted by General Maximus
but it also creates a single point of failure which has been demonstrated quite nicely over the last 3 weeks.
|
Not exactly. VM's DNS servers aren't a single point of failure, they're made of multiple redundant systems with active and passive backups. Including a minimum of two DNS servers being assigned by DHCP which gives, by definition, at least two different points of failure.
Course, that's how any sane network would be engineered. If VM's backend DNS services are actually designed around a single physical point of failure then that's some pretty epic fail by the network architects. But conversely, I'm pretty sure there's a dozen different parts of the DNS service that can blow up or die and you wouldn't notice a thing.
In the end any consumer service is going to have multiple single-point-of-failure chokepoints anyway, namely, the modem, the line coming into your house, your cabinet, and so forth. It's not meant to be uber reliable.