I am a user currently in Asia, and I’ve selected AP1 as my region. My connection with Storj is kind of slow, so I open another account using US1 as my region to test. I realized both of my account (AP1 and US1) both of my files, the majority of my pieces are stored in Europe.
AP1
US1
I like to know how does Storj decide what region to store my files?
and is there any way I can decide what region to store my files?
Thanks
To upload your uplink requests from the satellite 110 random nodes for each segment of the file, then your uplink starting uploads in parallel, when the first 80 are finished, all other got canceled.
To download your uplink requests 39 nodes from 80 containing pieces of each segment and start downloads in parallel, when the first 29 are finished, all remained got canceled.
This way you have a deal with closest nodes to your location.
By the way - did you use a native integration or S3-compatible?
hi Alexey,
So what’s the purpose and differences of region (AP1, US1, EU1) when I sign up if the files are uploaded worldwide ?
also how does the bandwidth usage calculation based on in the case of upload, is it based on the 110 connections attempted uploads or the 80 that’s completed?
say I upload a 110MB file that’s being uploaded to 110 nodes, assuming each segment is just 1mb, once upload to 80 nodes are completed, the rest are canceled. Assume all other 30 nodes are only uploaded 50%, my bandwidth usage are 80MB + (30*50%)MB= 95MB or just 80MB?
I’m using S3-compatible, currently only Rclone and Filezilla are natively integrated, correct? would it make a different in turn of what region my files will be upload.
Thanks
It’s a satellite, where your metadata is stored. To have less latency when your uplink working with metadata (listing for example), it’s better to use a geographically close region. But regarding data - it doesn’t matter too much, data is transferred directly between your uplink and the nodes.
Bandwidth usage calculated for egress (downloads from the network). You will be billed only for settled egress bandwidth (it will be confirmed by nodes and your uplink).
However, since you use a Storj-hosted S3-Compatible Gateway, this overusage is not affecting your limits.
But if you mean using local bandwidth in case of native connection, then yes, the bandwidth will be used from 1 to 110/80 times more than the amount of pieces. There is also expansion factor 80/29 for your actual size of the file during encryption and erasure coding.
1MB * 80/29 ~ 2.76MB
Then it will be uploaded to the nodes
2.76MB * 110/80 = 3.795MB
So bandwidth will be used from 2.76MB to 3.8MB during upload.
However, it’s actual only for a native integration.
For S3 there is no expansion factor on your side and no overusage of bandwidth, all this happening on the GatewayMT with usage of server-side encryption.
In any case you will be billed for storing the original size of your files and settled egress bandwidth.
There’s just not many nodes outside of Europe, so all satellites (AP1, US1, etc.) don’t really have much choice in this regard. See the map on the bottom of this page for the location of all active nodes: https://storjnet.info/
This might mean a business opportunity to a potential storage node operator located in east Asia (though, with the current geopolitical circumstances it might be quite significant which side of the Great Firewall of China the node is operating on).
The bill is based on settled egress. Thanks for clarification!
I also see some nodes in East Asia.
I’m based in Taiwan, and I’m surprised to see when I upload the file this morning, almost all the nodes are stored in Europe and 1-2nodes in US, but none in Asia region.
I also see some nodes in China, but I think it’s because Storj has not gain popularity yet, I don’t think it will take long before the GFW caught it.
The Asian nodes may simply be full, given they’re not as numerous. Though, only Storj knows whether this is true.
Could you please try to upload with a native integration?
are you talking about uplink CLI?
what options of software are there for native integration? (I currently known of Filezilla and Rclone)
any options for mobile ios?
thanks
Yes, I’m talking about uplink CLI or rclone or FileZilla, all other (except backup tools and Nextcloud) mentioned in our documentation uses GatewayMT.
From mobile apps I know only about this one: https://play.google.com/store/apps/details?id=com.storj_mobile
The Photos+ uses GatewayMT as well.
Got it!
I will give those a try, maybe I didn’t set up properly using S3 compatible access.
Thanks!