I’m pleased to announce that we’ll be rolling out some fresh Cloud infrastructure this coming Monday that will improve performance, security, uptime, and reliability for all tiers of the Spark Cloud (api.spark.io, device-service, cloud compiler, and webhooks). We’ll be making these changes between Monday 11am and 1pm CST.
Though we’ve tested things extensively and done everything we can to minimize the impact of this change, we definitely expect there to be a few minutes in that window when DNS changes are propogating where devices might struggle to connect or the API doesn’t behave quite right.
We’ll keep you posted of our progress and let you know where we’re all done on this thread.
This is just a quick reminder that this scheduled maintenance window is approaching, you may notice a few blips, but otherwise it should be very smooth. Yay upgrades!
Okay, we’ve completed the scheduled maintenance this morning, everything is running on shiny new hardware, and we’re very excited about it here! Bigger boxes, more redundancy, and better automation, woo!
Onwards and upwards!
Thanks,
David
(please feel free to report any weirdness you might experience here, we like to be cautious with changes, so we’ll keep watching things closely)
My friend and I are both working with Sparks. He lives about 20 miles from me. We have a Javascript application that talks to our cores. We each run that application in our own browsers in our own homes.
This morning (10am PST) and this afternoon (4pm PST) we have had intermittent problems. Our calls to the Spark Cloud to reach our cores just time out.
I know the first step is to question our Javascript or Core code, but we are working on other things right now and the code has not been updated. For example, I noticed this problem while working with Javascript in my laptop browser running a local web server. I thought it might be my problem so I went to the version hosted on my public web site. That cannot change by accident and hasn’t changed in several days.
Anyone have insight into why the spark cloud comes and goes these days?
I moved your post to this thread so that @Dave and @jgoggins would see it. The timing you are reporting is not quite right for the cloud update–my IFTTT messages say they rolled it out at 9:04AM PDT.
Are you still having problems? They should be all resolved from the upgrade this morning but you might have to reboot your cores and reconnect your web pages to the API.
Thanks for posting! We had a scheduled maintenance window this morning on the cloud and rolled out some massive infrastructure improvements. We try to give as much notice as possible before this kind of window, and generally downtimes are very small (under a second to a few minutes). Today’s downtime was slightly longer due to the nature of the changes, but that kind of downtime should be very infrequent.
Are you still seeing issues? If you want you can msg me your device ids and I can look into the issues you’re seeing, or you can email hello@spark.io and they can help you troubleshoot. At the moment things are looking normal from here, so please let us know if you’re still seeing issues.
