We are now a good ways along with restoring the database. All reference information was recovered, with special thanks to Alain LM for working with the partially corrupted database file.
Unfortunately we probably lost 10's of thousands of images. People rely on these images to research their collections and decide what they are interested in collecting. We have over one hundred contributors who work as volunteers to find and create photography to enhance the value of TroveStar to its users and we are sad to have lost any of their work. TroveStar is a labor of love for all of us: admins, curators and users and we hate to see the product of our hard work be destroyed.
However, we find it unacceptable that so many image files were lost. We used what we thought was a reputable ISP and we paid for a professional grade backup system to make nightly copies of our data and files.
These backups did not work. What was really frustrating was the lack of caring or effort on the part of our hosting service to help.
We needed a better solution for the future. We need to instill confidence in our contributors so they know when they add a photograph to our system it won't be lost. Fortunately there is an industrial strength data hosting platform out there that is both reliable and affordable. This is the Amazon Web Services platform.
We have already moved a large section of data over to this platform. What is especially important to know is that all new image files are stored directly onto Amazon. The Amazon platform is the same platform they used for their online retail platform and it is now the world standard for industrial strength cloud storage. And they make our old ISP look like the Keystone Kops.
So we want you to know that there will not be a 'Next Time' when it comes to recovering our data. We are pushing TroveStar onto a new, more resilient platform trusted by major corporations the world round. Please, add data with confidence your work won't be lost next time.
Thanks for your patience and understanding.