Hi Riley, That's a legitimate question, and one that we haven't officially answered yet. It's certainly the direction that things have been moving simply due to the nature of production management. Given that the SLA applies to HRD and not master/slave applications, you are definitely going to get a better quality of service migrating to HRD. In fact, I strongly advise that you do so.
One challenge that we have when dealing with issues is to decide whether we should do emergency maintenance that requires downtime. With any production system, it's not always guaranteed that maintenance will result in issues being completely resolved, which would be really bad for app developers. At what threshold do we determine that a downtime with no guarantee of addressing the issues is worthwhile? Global 0.1% error rate? 1%? The call is not always clear cut because those errors may not be evenly distributed, and the impact may be huge, or it may be small. With master/slave applications, we do what we can to address the short term symptoms as well as the underlying system issues without impacting serving, which is often an order of magnitude more difficult (It kind of reminds me of that scene in Indiana Jones where he takes an artifact, swapping it with a bag of sand as quickly as possible to try to avoid setting off traps. Pillaging of historic artifacts is way easier when it's not dangerous, not speaking from personal experience). When your application runs on High Replication, the call is easy: there's no downtime required in 99% of cases, so we perform the maintenance right away because if it doesn't address the issue, there's no serving downtime for users. If you're not subscribed to downtime-notify, I recommend that you do so. Announcements like this will NOT and never will be moving to StackOverflow: https://groups.google.com/forum/?fromgroups#!forum/google-appengine-downtime-notify We may be announcing a maintenance in the very near future that will impact the serving of master/slave applications. -- Ikai Lan Developer Programs Engineer, Google App Engine plus.ikailan.com On Mon, Mar 12, 2012 at 9:46 AM, Riley <[email protected]> wrote: > Ikai, it sounds like support for HR apps is being prioritized. Is that > the case? Should we expect that to be the case in the future? Sorry if > that's documented somewhere already~ > > Riley > > > On Monday, March 12, 2012 11:44:59 AM UTC-5, Riley wrote: >> >> Our appid is activegrade, we use the m/s datastore, and get from 0-10 QPS >> throughout the day. Normally we have 1-4 instances running, but since this >> seems *mostly* related to startup, we dedicated 10 idle resident instances >> to run all the time. This covers us a little, but still, when a user >> triggers a new instance, they get the 60+ second wait and then an error. >> Ugh! Our costs are relatively minor - about $20 a day now that we are >> running these 10 ordinarily unnecessary instances - but this is a big cost >> for us, and embarrassing too. >> >> On Tuesday, March 6, 2012 3:17:37 PM UTC-6, Adam Sherman wrote: >>> >>> Am I the only one seeing short duration outages? They are being >>> reflected at: >>> >>> http://code.google.com/status/**appengine<http://code.google.com/status/appengine> >>> >>> But I don't see anyone else complaining anywhere, so it makes me worried. >>> >>> A. >>> >>> -- > 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/-/VzKRK5UG96MJ. > > To post to this group, send email to [email protected]. > To unsubscribe from this group, send email to > [email protected]. > For more options, visit this group at > http://groups.google.com/group/google-appengine?hl=en. > -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
