Just for those of you that are interested.
I run two batch of tests with this, it worked very well as an application, done exacty what I wanted.
First test was run from our office from my laptop on a 3mbps down and 800kbps up connection to my FTP server at home (1ghz p3) load avegae was about 30-40% during a 50thread 10 image test (10 x 50 = 500 concurrent sessions sending these 10 files per folder). No server bottle neck however it reduced the upload down to 10kbps and crippled our upload and the test couldnt finish as the bandwidth just couldnt cope.
2nd test I run from our server in Telehouse north to the actual camera server which will be using, this is located in another datacentre up north. Exactly the same test, transfer time was much much better and test finished within 20 seconds, server load 3/4 and ram 1/2.
In reality, with 50 cameras, they wouldnt be sending across this many connections and transfers all in one go like it did in the test, so hopefully the server load would be a lot less that the test results from the concurrent mass testing.
All good
I run two batch of tests with this, it worked very well as an application, done exacty what I wanted.
First test was run from our office from my laptop on a 3mbps down and 800kbps up connection to my FTP server at home (1ghz p3) load avegae was about 30-40% during a 50thread 10 image test (10 x 50 = 500 concurrent sessions sending these 10 files per folder). No server bottle neck however it reduced the upload down to 10kbps and crippled our upload and the test couldnt finish as the bandwidth just couldnt cope.
2nd test I run from our server in Telehouse north to the actual camera server which will be using, this is located in another datacentre up north. Exactly the same test, transfer time was much much better and test finished within 20 seconds, server load 3/4 and ram 1/2.
In reality, with 50 cameras, they wouldnt be sending across this many connections and transfers all in one go like it did in the test, so hopefully the server load would be a lot less that the test results from the concurrent mass testing.
All good
