The App shows “no data” for the northern Peru 7.5 on 11-28 1052Z earthquake. I was very unhappy to think my shake failed during a major event, but looked and see it was at least captured locally, but not sent to the server. I could find no reason why, but looking at multiple stations around me, I see all but one that I monitor had the same problem.
Was the server down, and if so, why?
Hope you get an answer to this, as I was wondering why this has happened. I have noticed that it has happened several times around some major quakes, and I also never got an answer to a query I sent a while ago asking why none of the quakes in the Canary Islands appear in the events list. It almost seems like some events are vetted and/or removed deliberately, which is not good at all.
Same here, “No Data” for that timeframe and when I look at the Boom in Shakenet for that event, the waveform/frequency is blank up until around 11:04 UTC when it starts recording again.
Edit: When I look at the 24 hour plot for my Shake, it shows a big gap from about 09:47 UTC to about 12:02 UTC.
To all the Shake Community members:
I would like to offer my sincere apologies for the data loss yesterday, Sunday 28-Nov-21, for which I take full personal responsibility. Due to a confluence of unfortunate circumstances, the server ran into a problem which was unable to be remedied for the two-plus hours, during which data capture server-side was not possible.
We know that the community expects the network to have an up-time of at least 100%, and so do I. We, and the community, are very lucky with the technical team we have here at Raspberry Shake. Everyone, to a fault, is dedicated and committed to the cause, deeply believes in our mission, and works every day to ensure the integrity of the instruments and the network.
Rest assured, we have already undertaken measures to prevent this from happening again. Further, a new server architecture will be rolled out in Q1-2022 which will make these kinds of outages unable to happen. One way this will be done is to have built-in fail-over redundancy between multiple servers, as well as data back-fill capability from individual instruments to the server, neither of which exist at the moment. What this essentially means is that if the data has been recorded on your local instrument, it will eventually reside on the data server.
Until then, we will endeavor to understand all the details that contributed to this problem and will implement whatever methods necessary to prevent any recurrence.
We all thank you for your continued interest and support in the Shake products. We are still growing from our very humble beginnings and at various points suffer growing pains, as young companies do. You can trust, however, that we are continually learning from these to provide a more robust product that continues to improve.
please know that we are not removing any events from the event list for any reason whatsoever. it is true that the event list does not cover every region, no matter the size or remoteness, all over the world. this is a separate issue from the server outage which occurred yesterday.
the event list functionality is scheduled to be expanded in the next couple of months, events from the Canary Islands will specifically be looked at.
Thanks for the reply and update Richard. The store-and-forward data recovery sounds fantastic!