I don't know the details of the migration failure, but I know that prior to migration and prior to Rupert buying the site and the hardware it was all on a big E

I don't know the details of the migration failure, but I know that prior to migration and prior to Rupert buying the site and the hardware it was all on a big EBCDIC database with no de-duplication and I have no idea why this type of database was chosen. I did not keep in touch with the previous owners of the site. I did not work for them directly but managed the back-end of many big sites at the time. It was originally on a HP SureStore 256, then a HP SureStore 512 that I managed. I have no idea what happened after all the hardware left our data-center. It was never meant to be a social media site but rather a mountable drive on peoples laptops part of another acquisition and somehow evolved into a social media platform. The SureStore is just a rebranded Hitachi HDS and one drive failing will not cause loss. In fact one would have to try hard to lose data on those storage arrays via bad disks. If I had to speculate, it could have been something to do with migrating a 19TB EBCDIC database, but that's just a guess. Some prior discussion [1] [1] - https://news.ycombinator.com/item?id=39058953 https://news.ycombinator.com/item?id=39058953