HN user

jcanyc

10 karma
Posts0
Comments11
View on HN
No posts found.

In terms of cost/expense, AWS is very cost effective for this type of work load.

If you build out the skill set and scripts to run this type of job, you can process the images, store them on S3 and kill the running infrastructure. You can then run this process for big and small workloads as the needs arise on demand without having to own servers or space in a data center.

First consider thumbnailing these in the fly as they're requested by the user. But if you must mass convert them, this job must be parallelized among many compute nodes; overly tuning the processing of each conversion probably won't be very fruitful.

You probably want to move these images to AWS S3 and run compute jobs and upload the resulting images back up to S3. You could create AWS Simple Queue Messages with the S3 URLs of each of the images and pop messages and autoscale EC2 instances based on the depth of that queue. What's the plan for these files after they're processed?

I am local, have deep AWS experience and have next week off if you'd like some pro bono advice.

50TB S3 ~$1500/month SQS ~$0.50/million messages

The cheapest Github account is extremely economical if you price out all the features you get with the account. For your needs and price point, I don't think you're going to find (or even be able to identify) much of a difference for from your end users' experience.

share.php?r=66a5813d my session.

I wasn't paying attention and clicked the Play button and thought I was about to see a presentation. Imagine my surprise as a professional *nix systems admin as I was presented with a root terminal open on a gnome session. I didn't realize how new the posting was and I was caught off guard and flabbergasted.

With the haste of a thief I moused over to the terminal and ran curl icanhazip.com to get the external IP. As I was thinking to myself, "This must be someone's fuck up", I started a basic port scan on the external ip from a another workstation. Seeing it was running, and back on the pinemango.com vm I telnetted to a port in the low 5000s I figured was vnc, then the session crapped out. Right, I've seen it do that before (whoops).

Cool idea, I've inferred some of the stack from my side of things. I'd love to know your thoughts or war stories about managing the systems.

Why an old school version of Ubuntu, just wondering? Would you like help scripting/pairing down the user session?