Open forki opened 8 years ago
In production, it is currently using Azure blob storage - what is in the local data directory is basically mirrored in blob storage. The code for this is in the azure.fs file. The site picks Azure or Local file system here depending on an environment variable (set on Azure).
I went with blob storage/file system because it is the most basic thing that would work (we would have to store the metadata index in a database if we had multiple machines hosting this, but given the site load, that seems overkill :))
Thx. Just for fun I will investigate if I can come up with a docker solution. On Dec 3, 2015 22:53, "Tomas Petricek" notifications@github.com wrote:
In production, it is currently using Azure blob storage - what is in the local data directory is basically mirrored in blob storage. The code for this is in the azure.fs file https://github.com/tpetricek/FsSnip.Website/blob/master/code/common/storage/azure.fs. The site picks Azure or Local file system here https://github.com/tpetricek/FsSnip.Website/blob/master/code/common/data.fs#L19 depending on an environment variable (set on Azure).
I went with blob storage/file system because it is the most basic thing that would work (we would have to store the metadata index in a database if we had multiple machines hosting this, but given the site load, that seems overkill :))
— Reply to this email directly or view it on GitHub https://github.com/tpetricek/FsSnip.Website/issues/44#issuecomment-161796689 .
If there is something similar that we could use on Amazon, that would be pretty epic. Would love to be able to host this for cheap on AWS :-)
Perhaps @adron would know something?
@tpetricek @forki AWS is having RDS. I don't know the price but they do have many database options.
Just for curiosity why not using file based database. Store locally on cheap EC2 or Digital ocean server.
Or may be localjson store. @robconery worked on something long back.
I was talking about something like biggy. have a look may be it will solve Database problem.
What is the database story?
Currently we need to download a zip to get started. How is this working in production?