2 ms·
Time is one factor, it's also surprisingly easy to corrupt vfx master files with just a few missing bits. Checksums will identify something went wrong, and the
by diveanon 6y ago
Time is one factor, it's also surprisingly easy to corrupt vfx master files with just a few missing bits.
Checksums will identify something went wrong, and then you need to redownload the file to a quarantine network and scan it. Takes time.
Much easier to go from trusted source to trusted source and verify the files on drive prior to shipping.
Amazon just released a device (forget the name) for exactly the same use case. We developed it in house.
Not to mention most major studios will contractually prevent you from exposing anything to the web.
- itismetheidiot 6y agoFYR Amazon has the AWS Snow Family with the largest being a 45ft TEU, while Microsoft does Azure Data Boxes that are somewhat smaller.
- closeparen 6y agoApologies for all the questions, I'm just curious about this. >Checksums will identify something went wrong, and then you need to redownload the file to a quarantine network and scan it. Takes time. Surely a sensible file transfer algorithm would compute checksums on small and easy-to-retransmit chunks? Does rsync not do this? Isn't it already happening in TCP? >Not to mention most major studios will contractually prevent you from exposing anything to the web. I understand that workstations with media on them are not going to have internet access, but do they really prohibit site to site VPNs? >Amazon just released a device (forget the name) for exactly the same use case. We developed it in house. Snowball and Snowmobile? IIRC these are primarily meant for one-time migration from on-prem storage to S3. Do people really use them on an ongoing basis?