Hello,

I downloaded, installed, and began using your program and things seemed to be working as expected.

The bucket that I was working with has a fairly large number of files (my server backs of a database every 30 minutes and copies the backup to the S3 bucket).

After downloading a few files (5) and then deleting quite a few (216), my computer seemed to slow to a crawl and the program would take about 10 seconds to respond to a click (e.g. right clicking and the menu wouldn't come up for roughly 10 seconds).

I opened task manager and noticed that the javaw.exe was using 416MB of RAM. (my machine has 1.5GB of RAM).

I closed the program down and this took care of it.

I re-opened the program and connected to the same bucket. I selected roughly 50 files to download and began downloading them.

The javaw.exe file began climbing from 250MB of RAM and while downloading was always using between 75%-100% of the CPU.

By the time the files were downloaded, the memory usage was at 400MB and climbing.

I let the system sit idle and the memory usage continues to increase, roughly 1MB per second. CPU usage remains high as well.

Once I close the program, the javaw.exe app is removed and the CPU usage returns to normal (0-15%).

The latest version of java for my OS (windows XP) was installed within the last few days.

On another note, a nice feature would be the ability to filter the file listing to only show files that match a particular pattern.

Thanks,

Joe

Joe,

Thanks for the feedback. We have made several improvements recently for the large buckets. You can download the sept beta (1st link in this forum), which has paging and chunking, i.e. we will only display x number of files per page, which should improve the memory issues and load time. The default page size is 1,000 objects, which can be modified via - Tools - Preferences.

The prefix option is a good one and we will add it in the next release.

Again, please do check this forum by next Tuesday for an updated version because most of the changes in past 6-7 weeks are related to performance on large buckets.

Thanks
Saurabh

Filter by prefix was added in October Released. I am not closing this thread as "solved", and am moving it to deferred feature request forum, as we would like to perform some additional tests and optimize it more for CPU & memory when dealing with large buckets.