Do maintentance periods really have to be during core working hours?

[Per]

Although I'm overall happy with GAE, and I know that the maintenance periods "only" affect M/S. And I don't mind the occasional one either. But I find it increasingly frustrating that they always have to be scheduled in the middle of the week, and during a time that most of our US clients would still call core working hours.

I am now pasting the third downtime notice in a month onto our application's dashboard. Yeah I know that one maintenance period didn't actually happen, but it was cancelled too late, all our clients had already gotten our downtime warning, which is almost as bad.

I am wondering why these maintenance periods cannot be done on the weekend, or at least later during the day? How about 6pm to 8pm PST for instance? That's still late at night in Europe, it's just about as annoying for APAC, but at least it's a lot better for US users. And it's not *that* late so that East Coast Googlers have to get up in the dead of the night either.

Or am mistaken and the typical usage pattern is "after hours", because nobody except for us dared to use GAE for an enterprise application? :)

[Ikai Lan (Google)]

Our next maintenance period is from 6-8pm, but that's a coincidence more than any other reason.

First off, we really want to encourage you to migrate to High Replication. I can't stress this enough. This allows us to do lots of work to App Engine without impacting your application's service. I know some people have been holding off because the blobstore migration tool doesn't exist yet; this is coming.

One of the reasons we like having maintenances at certain times is because of one of App Engine's strengths: it's built using a lot of Google's existing infrastructure (the datastore uses Megastore, which uses BigTable, for instance). By performing maintenance during US/Pacific working hours (or close), we can always ensure that if we need to contact another team, that we can minimize the delay of finding someone at home.

We love High Replication because it lets us hide away all of the work that needs to happen to keep a system as large as App Engine online. Please embrace it!

--

Ikai Lan 
Developer Programs Engineer, Google App Engine

plus.ikailan.com

--
You received this message because you are subscribed to the Google Groups "Google App Engine" group.
To view this discussion on the web visit https://groups.google.com/d/msg/google-appengine/-/bK7kFbD_8E4J.
To post to this group, send email to google-a...@googlegroups.com.
To unsubscribe from this group, send email to google-appengi...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.

[Greg]

Just a reminder that "core working hours" for many of us are different
from yours. Back when we were on MS, it seemed like every outage was
in New Zealand working hours. Basically Google get it in the neck from
somewhere around the globe no matter when they schedule their
maintenance.

And the real answer to your problem is... HR. If you care about your
app enough to complain about outages, you should make migrating to HR
your top priority. It will change your life, really - no more clenched
guts when you get an email about the next maintenance period, or
bitten fingernails when they go over the scheduled time. Plus you get
to make snide comments to MS-users like this... :)

Cheers!
Greg.

[Brandon Wirtz]

I chose M/S because even more than a mermaid, I want a girl in a Dominatrix
outfit.

(I got the pool, so I'm one step closer on the Mermaid thing)

[Greg]

As usual, Brandon, I've got to tip my hat to you - you've come up with
the only good reason to choose MS over HR.