I will do that when I get home in a few hours.

jscher2000,

  No, I arrive on a page of results and the page itself may have a hundred links of results. This memory climb happens as I click on the links, go back and click on the next link. 

For what I mean by the Application cache, I will lead you to it.

Go to Preferences/Options | Advanced | Network and on this tab there are two caches listed: "Cached Web Content" and Offline Web Content and User Data"

The latter has a line that reads:

"Your application cache is currently using 0 bytes of disk space". However, when it runs into this huge memory consumption problem, you can only get FF to release memory by clicking "Clear Now" on BOTH of these caches. So, it is my suggestion that the "application cache" may be misreporting the memory it holds. I have no internal evidence of that--it's just a supposition.

Noah_SUMO

Thanks for joining in and getting dev assistance. Off topic but just a thought, would it help if this forum was able to handle, Gzipped json attachments. That is how later versions of Firefox export about:memory results. https://blog.mozilla.org/nnethercote/2013/04/10/gzipped-json-is-now-the-preferred-format-for-attaching-memory-report-data-to-bugs/

They also now import and even diff the results.

curmudgeon99

By now you will be seeing about:memory data. You may also be interested to note there are also

• about:compartments
• about:cache that then splits off giving further info including offline cache
  • It gives the location of the offline cache, mine is on the HDD not in RAM

Okay, gentlemen, I have performed the test as you requested.

This was with FF 21.0 because the problem was worse with 22.0.

The begin memory details are here: http://pastebin.com/aAJL7C1g = BEGIN

The during lockup memory settings were much, much harder to get... because FF was locked up. It started to lock up at 2.31GB and while I was trying to get it to respond to a click in the address bar, it ranged from 2.80GB to 2.31, which is where it was at about 15 minutes after the lockup when I was able to get this memory. http://pastebin.com/V2kKNMei = Freeze

So, if this shows the memory at less than the 2.90GB I mentioned, that's because it was impossible to get FF to give me a memory dump at that level. Hope this helps.

Crazy result, no wonder it freezes.

• 2,129.98 MB (80.69%) ── heap-unclassified
• with loads of tabs open and running for hours I have
  • ─81.78 MB (15.06%) ── heap-unclassified

Just a thought:
Now you have seen about:memory you could try using the buttons on it to reduce memory usage. It may be very inconvenient but if it temporarily clears the memory it should help you un-freeze without re-starting.

John99, That was ONE tab. So, I am gratified that you guys are seeing objective confirmation of the problem I've been dealing with.

As I wrote earlier, the only way I can keep FF usable is to ride the Clear Cache and Clear App cache buttons so I don't need to do that test again. But honestly I use FF as little as possible now since Chrome shows trivial memory usage at all times. I hope for the good of all FF users your devs can use this data to figure out where the leak is happening.

I doubt that test alone will be sufficient for the developers. They may have other questions and tests they would like running. This may well be an issue that they are not yet able to replicate.

An obvious question that no one seems to have asked you yet: Does the problem only occur when a certain site is used, or do you get this memory leak occurring on any and all sites that you have tried.

For instance if you had this forum as your one open tab and browsed the forum for 10-15 minutes would that freeze the machine ?

I do not use a Mac and so do not understand its activity monitor. I was pointing out that the Firefox about memory not only gives dynamic results it also has buttons; in my case at the top of the screen; that allow memory use to be reduced.

I don't know if this happens on other sites because I have taken to the habit of killing tabs frequently. However, this site, which is from the State of Indiana, is a for pay site and I use it to search legal records. There is no alternative source for this information and a search of even one day's duration (shortest possible) can turn up 600 links. So, there really is no feasible way to avoid using it in the manner I am.

So, in my 16 years as a software developer I have noticed a phenomenon where developers look for any excuse to sidestep the obvious conclusion that their code has a problem. And however many times that process is followed, the original defect turns up to have been valid. Every single time, in my experience. So, I don't want to get into a situation of "blaming the victim". Are you seriously arguing that the website is at fault? I have always believe that your code should not be that vulnerable that any HTML site can take it down. I would better say that the use case of this site has exposed a memory leak in FF. Also, from reading this thread, I am not the only person who has faced this, so I really don't think it's going to be productive to blame the victim here, which is where your comment is heading.

I have given you the memory dumps that were requested--validating my problem. The Mac Activity Monitor is very similar to what you would see in the Windows Task Manager. It reflects memory consumption. It exactly matched what your Firefox about:memory reported so I don't think that's germane to the issue.

So, I will of course not mark this as solved. I will consider it abandoned along with Firefox. And you can blame the website I was using and the website the other reporter was using until more people complain and then eventually some developer will find the original defect and I will not hear boo about it.

Thank you for trying. It's not worth using FF if I have to babysit its memory consumption.

Hi curmudgeon99, no one is blaming you for anything.

Surely you recognize the possibility that each site's scripts and use of plugins are different. The developers and QA team and alpha and beta testers certainly can't have seen everything. The question is whether they should focus in on this site to better understand the problem, or whether 10 minutes of heavy browsing on other sites causes you the same grief, in which case the site isn't especially relevant.

On the contrary, I have always followed the practice as a developer that my applications should be made bulletproof. I fail to see how any use of JavaScript should cause a browser to keep 2.9GB of RAM. It does not seem feasible--and now we know those numbers are actual.

I cannot imagine explaining to the business customers that I work for year in year out that our application will work well as long as some stray JavaScript doesn't take it down. It just sounds unprofessional and I cannot imagine any way a rogue site should expose that vulnerability. Chrome does not experience this problem on the same site, as I have discovered since I switched to it. It's not the site's problem--it's firefox. FF has had a memory leak since version 4.0.

Try an experiment: Google this: Firefox "memory leak"

It turns up 468,000 hits. Are you saying all those people are using this one site in Indiana? There is a person ON THIS SAME THREAD who is seeing a memory leak somewhere else.

Hi curmudgeon99, over 1400 bugs have been filed in Mozilla's bug tracker that contain the phrase "memory leak" and Firefox somewhere in the record:

site:bugzilla.mozilla.org/show_bug.cgi "memory leak" firefox - Google Search

Flaws exist in Firefox, no doubt about it. The question is whether the site you use day in and day out has surfaced something unique or not, or whether your configuration is subject to this excessive memory use on lots of sites. Those would be two very different problems.

Thank you for your time.

initial edit -now with additional links will tidy up later today have other work to do now

Thanks a lot for your help curmudgeon99

I realise you will be busy and so not be able to give time helping out with this. I also empathise with the view the Firefox ideally should be perfect and not have any hard to track-down edge case problems. That is especially the case if Firefox is used with no extensions. Something is seriously ^{A} amiss here.

I have linked potentially related information with some links appended to the bottom of the post and in ^{little super-script}

ANYONE ELSE WANTING TO HELP ?

Note, in the unlikely event that we get more responses that we need. I will edit this post to let you know.

If this really is a problem affecting lots of people can you be someone that assists in helping them by helping us investigate.

If anyone else reading this feels they see a very similar problem of

1. using a Mac
2. in Firefox's safe mode ^[SM]
3. and with all add-ons disabled
4. running up a very high memory Usage, in a matter of minutes
5. crashing, or hanging; once memory is high
6. when surfing around using just one open tab, or on a session staying within one particular site

Then initially; in my view; this is what may be helpful to us trying to understand the problem better so that we are able to reproduce it and so enable Firefox developers to look into the problem and attempt to solve it for all of us.

1. Please note at the moment I am initially specifically looking for very similar cases ideally where all the above apply.
2. Make your views known
3. Start your own thread
4. Provide information
5. Take note of the difficulties involved

1.) Similar Cases
The listing above is my suggestions about factors needed to consider a problem to be similar or strongly related. Please feel free to post; elsewhere in the forum; as usual, about any memory etc related problems, but in this thread lets try to stick to the subject in the title.

2.) Make your views known
All of this helps us to understand the importance of a problem

• View - just visiting and viewing the thread is logged, note the number of views is displayed
• Vote - if applicable
  - Use the me-too button [I have this, problem too]
  - Use the [Helpful button] (? I think that may only allows a logged in user to vote on one post)
• Post Feedback - See ^[F]

3.) Start Your Own Thread
it is impossible to follow threads that have multiple individual problems in them. Often the problems have a similar symptom, but may have multiple or differing causes or solutions.

• This really is essential please use /questions/newspecial direct quick link.
  • Please try to follow the prompts to add the troubleshooting information when you do so. Please mention in your new post if there were any problems trying to do this. (It should for instance make your list of add-ons available in the thread).
• Post a comment back here in the current thread, just to confirm you have posted a new thread. Others will be able to find and read your post.

4.) Provide Information
Once you have started a new thread please explain how you see your problem and describe briefly what you are doing as this occurs.

• If you see any pattern and particularly if you know of anything that seems to cause, or seems to prevent that then please comment.
• We really would like detailed instructions on how create and observe this problem. For whatever reason it probably only affects a small minority %wise of users.
• Run the about:memory test ^{T} and post the information in your thread with links to your pastebin ^{P} If you need clarification. Or want to run additional tests please comment in your own thread.

5.) Difficulties
Sorry if you have a problem. Firefox is always evolving, and trying to improve. I know they may not always get things right first time, and know some problems get neglected, but as users can we work together to help any dev trying to collect information on this problem.

Saying also have a problem may sometimes help, especially if you use the methods I suggested in this post above. On this forum generally we try to keep a single problem to in a single thread as in the ^[G] Guidelines. Multiple problems just clutter the thread and make it hard to follow. Comments should be proposed solutions or questions to obtain information in order to assist in solving the problem. However as I have already said feel free to comment in order to cross link to related threads.

if you wish to help you will need the spare time to run the test suggested, and it may then help; if necessary, to provide further information or run additional tests.

Curmudgeon99 reports this is an ongoing problem seen over several versions, but for whatever reason Firefox testing and reporting does not seem to have picked up any reproducible problem despite its severity.

LINKS

• [A] Agreement it is likely to be a real problem /questions/962986?page=2#answer-457036 (see also above) and
• [F] More information including Feedback Report & Results. Use Submission Link
• [G] Mozilla Support rules and guidelines suggest not asking a separate question in an existing thread.
• [P] pastebin, the site will handle code and other pastes that will not work well on our forum http://pastebin.com/
• Start a new thread use /questions/new
• [SM] Diagnose Firefox issues using Troubleshoot Mode
• [T] The initial test of running about:memory and collecting info to paste into pastebin is given in a post above /questions/962986#answer-456898

N.B.

• Noah, please feel free to edit this post as required.
• Jeff. , similar, just pm anyone with edit privileges, or get them yourself !
• I will link this from Contribs forum, maybe we have someone who can reproduce or find users who can. (Contribs Posts)

I noticed that it has gotten quite worse under FIrefox 23.0.. I would have one tab open (showing Mozilla Support page), and it would consume over 1Gb of memory. I have three main Add-Ons that I use, AdBlock Plus, DownThemAll, and "Flash and Video Download"... By disabling the later two, it helped reducing the memory foot print by maybe 100 to 150 Mb or so. That's not to say AdBlock Plus is causing the memory leak, because memory leak does not always happen right away (always after some time of usage).

Resetting Firefox did not seem to make much of a difference. However, it made me establish the sync again (ugh.. can't Firefox just maintain the sync relationship while resetting profile?)

Obviously that does not necessarily sound good,or expected. It may be interesting to see what you are doing to get Firefox to use 1GB of memory with only one tab open, but please start a new thread. On this occasion please the special direct link

• https://support.mozilla.org/en-US/questions/new/desktop/fix-problems/form?search=Firefox+suspected+memory+leak+%28see+also+questions%2F962986%29&step=aaq-question

You may wish to post back here to confirm you have posted a new question.

This thread is closed and archived, but if anyone should come across it note a Firefox bug causing Mac memory problem has been fixed in Firefox 28.

• see Firefox 26 on OS X 10.9.1 still has huge memory usuage (leak) which leads to crash /questions/981674?page=2#answer-539109
  • Bug 937878 - Extremely high layer memory usage on Mac when browser windows are minimized