Again about smugmug availability

lexunus

Both smugmug.com [63.81.134.23] and dgrin.com [63.81.134.25] were unavailable recently, with tracert stopping at
smugmug-gw.customer.alter.net [157.130.214.214]

Very disappointing.

Andy

lexunus wrote:

Both smugmug.com [63.81.134.23] and dgrin.com [63.81.134.25] were unavailable recently, with tracert stopping at
smugmug-gw.customer.alter.net [157.130.214.214]

Very disappointing.

Hi Lexunus,

Thanks for posting. Could you be more specific as to the exact time of the outages you experienced?

EDIT: OK sorry - I'm on the east coast of US so I didn't experience it last night. I will ask our Director of IT (Wireless) to comment on the reasons - it was Data Center maintenance, beyond our control FYI... (not that that helps much, down is down, right?)

We're sorry for any inconvenience that this caused.

jfriend

Something for you folks at Smugmug to think about...

Andy wrote:

EDIT: OK sorry - I'm on the east coast of US so I didn't experience it last night. I will ask our Director of IT (Wireless) to comment on the reasons - it was Data Center maintenance, beyond our control FYI... (not that that helps much, down is down, right?)

We're sorry for any inconvenience that this caused.

If dgrin.com and smugmug.com are hosted in the same datacenter and/or same cages, then when you have a smugmug.com issue, you may also lose the ability to communicate with your customers about it via dgrin (it depends upon the type of outage). However, if dgrin was hosted elsewhere with a different provider, then you are more likely to be able to communicate with your customers during a smugmug outage. Just an idea to consider as you grow.

Mike Lane

jfriend wrote:

If dgrin.com and smugmug.com are hosted in the same datacenter and/or same cages, then when you have a smugmug.com issue, you may also lose the ability to communicate with your customers about it via dgrin (it depends upon the type of outage). However, if dgrin was hosted elsewhere with a different provider, then you are more likely to be able to communicate with your customers during a smugmug outage. Just an idea to consider as you grow.

I can't think of any time that dgrin and smugmug were both down at the same time...

jfriend

Lots of reasons to think about this

Mike Lane wrote:

I can't think of any time that dgrin and smugmug were both down at the same time...

See the first post in this thread for a user who couldn't get to both at the same time. But, it doesn't actually matter whether it has happened in the past or not. I'm discussing the kinds of things you do when you are trying to maximize your chances of being able to communicate with customers during an unplanned service outage.

Whether or not dgrin was affected during a smugmug outage would obviously depend upon the type of outage. If it was a smugmug software problem or software maintenance issue or a specific server failure, then dgrin will probably work fine during the smugmug outage.

If it's a problem with the network provider or the ISP routers or a cage power problem or a cut fiber or a regional internet outage or a flood or a tornado or an earthquake or a hurricane, then diversity of geography and service providers for different parts of the service and ways of communicating with customers that don't rely on the same internet infrastructure as the core service is considered a wise move. It's something companies start to think about as they grow and the consequences of an unplanned outage and/or severed customer communications get greater.

Wireless

We're very sorry that there were connectivity issues last night. It appears there were two events that occured: scheduled maintenance with our colo provider, and a fiber cut . While we have multiple connections to our two datacenters, add the two events together and we get a complete loss of connectivity for both dgrin and smugmug.

Downtime is still downtime, whether it was our fault or our vendors. We CAN do better, and we will. It's going to take a bit of time, but we're working on adding yet another datacenter, more redundant network connections, and of course, additional storage!