Thursday, May 5, 2016

An Intruder Add-On to Firefox?

I periodically have severe slowing of my PC that cuts Internet transfer rates in half or more.  It goes away on reboot or doing the Firefox reset.  Nothing surprising shows up in the Task Manager.  It might be Flash Player, something that I uninstall when I am not using it with createspace.com.  But why does it take a while to start slowing thinks down?

Yes, it does look like a memory leak. When I started Firefox, and opened bloger.com, it needed about 200,000 KB.

Now, after opening several tabs and loading various web pages then closing those other tabs: 233,000 KB.   Amazing how fast thrashing VM can slow down your PC.

Chrome seems to be a more efficient user of RAM; so no surprise it's faster: 61,000 KB where Firefox was about 200,000 KB.

Also, speedof.me now reports better than 10 Mbps download.  Thanks all.  Curiously, the longer Chrome runs, the more instances appear in Task Manager, and downloads slow down.  Restart Chrome, performance comes back.

4 comments:

  1. Probably not the reason, but Firefox has horrible memory leaks, and a periodic restart is a good call.

    (I mean, if it makes you hit swap, that'd certainly explain it, but I'd hope you'd notice that specifically.)

    (re. Flash, might I suggest installing Chrome, which has a built-in self-updated Flash runtime, and simply using Chrome for createspace, if it supports it?

    That way you don't have to install/uninstall Flash over and over.)

    ReplyDelete
  2. I was just reading a few days ago about how firefox seems to naturally get slower with use, and the "reset firefox" feature is the Mozilla recommended prophylactic fix.

    ReplyDelete
  3. When you first open Firefox, open taskmgr and see how much RAM Firefox is using. After it has been open for a day or two or three, check the RAM usage again. I periodically close Firefox and re-open just to reclaim RAM.

    ReplyDelete
  4. Memory leaks: what a 20th century problem. In the late 1980s, I became so frustrated with malloc/free problems that I wrote a malloc/free pair that logged every call, then at exit, processed the log to see who was doing mallocs without a corresponding free. It worked so beautifully that I could not imagine that everyone wasn't doing that.

    Trying above suggestions including Chrome.

    ReplyDelete