Fastmail was Down and now everyone is squabbling...

[Sherry]

Originally Posted by sadams

Sherry,

I did post one - you can see some of the tweets at the time via the google "realtime" search for the period... this link ought to work:

Ok, most posts are wanting a "status update". Sounds like they know to check the status blog. If that's true the main problem with FM's communication would be them wanting more info in the blog so if FM improves their communications through the blog then those on twitter would keep up with what goes on without the need for FM to post to twitter (and other places like facebook) also?

Sherry

[Shelded]

+1 Sherry!

[sflorack]

Originally Posted by sadams

I see value (to FM/Opera and customers) in providing some info and reassurance during an outage incident.. I appreciate others don't value this and I respect their right to, and admire their patience in, calmly waiting it out :-)

I think users appreciate information regarding outages; what seems to differ is the amount of information you're personally expecting. I'd prefer they put all of their resources in fixing the problem, rather than giving a nerdy explanation most of us probably wouldn't understand anyways. I find a "all services are down, we're investigating" message to be appropriate, as it acknowledges they're aware of the issue.

[sadams]

Sherry,
A single authoritative source would be fine.
you could then rely on "the community" (or perhaps, rather "the communities" on various channels to pass things on).
However, it would be very straightforward to populate/update multiple channels via a single blog post.
Twitterfeed (twttierfeed.com) takes RSS feeds (say from a blog post) and auto-magically creates a headline and short-url-link to the orig blog item and posts it to a twitter account.
*Note there is a delay/wait whilst this happens - the service sleeps and re-check every 15 mins or so I think - that would need to be considered.
Various similar tools/apps can do that for Facebook too.

I also understand that various blogging tools/platforms have multiple options (such as plugins for Wordpress for example, or posterous.com/) which enable auto posting of a blogpost to other sites/services. *Not suggesting these are or should be used by FM/Opera - just giving examples.

The point being.. this stuff is becoming "normal" now... and I think for many (well, me anyway!) the absence of FM from these services is unnecessary and easily, inexpensively fixed.

Also - note that Bron posted that he was temporarily locked out of the status account during the outage (he gives the reason a few pages back in this thread). So - the lack of updates to the status blog were partly by choice (i.e. they didn't intend to post anything anyway) and partly through the process not taking into account likely knock-on effects of the outage meaning access wasn't available.

Cheers, Steve.

[sadams]

sflorack,
This has been discussed at some length in this thread.
Regular, brief updates give reassurance and helps customers know you're working on it.
Setting expectations can be really helpful - knowing when a service is likely to come back (once an ETA can realistically be determined) - lets people make alternative plans.

One guy gave the example of whether to switch over his business users to a backup service (or something) I think it required the switching of MX records (or something!).
You wouldn't want to do that unless the outage was likely to be for maybe a few hours.

So - good, authoritative and timely information can be really useful for some.
Yes it can be "hard to do" - but it has value.

For those who suggest - "it impacts the time to fix by distracting the techies" - I think this misunderstands what's required of the techies.
i.e. Brief, to the point updates.
Whilst I recognise giving regular updates doesn't get the system up any quicker - it shouldn't have any significant impact of the techie fix activity.

Cheers, Steve.

[Shelded]

@sadams I'm on your side WRT the need to update with something specific. OTOH I don't think you, a master of language, know how disruptive and time-consuming some of us find the task. Writing a blog entry for this type of event needs to be commercially acceptable, accurate, sensitive, all that stuff. Which a techie guy like me has to totally concentrate on. For too freaking long. And I never have an idea how long the unique train wreck will take to untangle. I would love to gripe at someone else on my team and, in turn, have them translate for the public consumption.

[Edwin]

Ok, while there has been some interesting discussion here, there have also been a lot of issues surrounding this thread and since it refers to an incident which itself was resolved several weeks ago, I have decided to close it here.