I have a strange memory issue I'm having problems resolving and would appreciate some advice as to where else to look.
The program I have (iPhone App) has a function whereby it basically downloads loads of files, processes those that are JSON, and stores the rest to disk. The JSON processing is CPU intensive and can take several seconds per file, so I have a NSOperationQueue with maxConcurrency limited to 1 that handles all the heavy lifting, and a queue that manages the multiple files to download.
Ever since iOS5 came out, the App has had problems completing the download sequence without crashing and so far what I have tried is;
1) Changed the performSelectorOnBackgroundThread JSON processing to use a single NSOperationQueue so as to limit the number of background threads working with large objects.
2) Added NSAutoReleasePools inside loops that create multiple, large, transient objects.
3) Flushed the sharedURLCache to ensure the files aren't hanging around in the system cache.
4) Stored the JSON objects to disk using NSKeyedArchiver and passed the filenames between threads rather than the actual objects, to again try to mitigate the number and size of retained objects currently in use.
All of these at first seemed to make a difference, and when I look at the memory allocations, I've now got the peak usage down from just over 20MB (hence no wonder it was crashing) to under 10MB, and yet the app is still crashing with low memory as before.
I'm trying to trace what is eating the memory causing the app to crash and on this occasion I'm having real problems persuading Instruments to tell me anything useful.
Here's a typical trace (on an iPhone 3GS running iOS 4.3.5)
You can see that the PEAK usage was a tad over 7MB and yet shortly after, you can see the 2 flags pertaining to low memory, and then low memory urgent, followed by the app terminating shortly thereafter.
If I use the memory monitor, the cause of the crash seems clear enough - physical memory is being exhausted - look at the light green trace below. The low memory warnings co-incide (not surprisingly) with the physical memory running out.
There are no leaks showing FWIW either (I've done that in other runs).
It's not image caches or NSURLConnection caches and the only thing I can think of is that perhaps there are some bad leaks that aren't being detected ... but I'm having issues identifying them because if I click into all allocations to see the objects that are live, and then do a command-A to select them all (in order to paste them into a spreadsheet to see where the memory seems to be), at the point I hit command-C to copy them, Instruments beachballs and never recovers.
I really cant figure out what's going on. Does anyone have some advice on how to persuade instruments to show me some more useful information about what is using this memory?
Sorry I can't post any meaningful code fragments ... hopefully the instruments screenshots at least give you an idea about where I'm coming from.
The Leaks instrument isn't terribly useful for figuring out anything but the obvious leaks in your app.
What you are describing is an ideal candidate for heapshot analysis.
tl;dr Heapshot analysis allows you to see exactly how the heap of your application grows between any two points of time (where you determine the points).
Related
This question already has answers here:
Problems with memory management, autorelease, permanent heap is sometimes 250+ kb on iOS
(2 answers)
Closed 9 years ago.
Hello my question is maybe General I am not asking for code etc.
I developing only for iPhones with iOS6.1 and above
When I run my application the RAM it uses only grows up(when I switching between views (I have like 15 views)).
However after I ran test with analyzer it didn't find any leaks.
also no leaks were found in instrument leaks.
Despite my application doesn't exceed 20 mb of RAM I am still worried that something may be just not ok there.
I am using ARC ,but the ram still goes up.
Is there any way I can check what can cause 1 sided ram allocation ?
If the memory continues to go up, it could be a variety of different things, but "strong reference cycle" is the prime suspect. Sadly, this won't necessarily show up in Leaks tool in Instruments, either.
Do snapshots/generations in the Allocations tool and identify what's not getting released (notably if it consists of any of your classes) and go from there. Specifically, run the app through its paces, then mark snapshot/generation, do a bit more, and then mark another snapshot/generation. Look at that second snapshot and see what's been allocated (but not released) captured since the prior snapshot, with a focus on your classes. You'll find the culprit pretty quickly that way.
See WWDC video iOS App Performance: Memory for practical demonstration.
For example, here is a healthy app that I profiled through the Instruments' "Leaks" tool, but I'm going to focus on the "Allocations" tool:
In this profile, I waited for the app to quiet down, tapped on "Mark Generation" button (resulting with "Generation A", the first flag in my timeline). I then went to a view and then dismissed it, and did "Mark Generation" again, getting "Generation B". The "Growth" column is telling me that between Generation A and B, 100kb was consumed, but not released. But I'm not worried about this yet because there might be some iOS internal cache of UIKit elements. So, I repeat this process one more time to get "Generation C". Now that's interesting, now reporting a growth of only 8.26kb, which is negligible. This, combined with a clean bill of health from the Leaks instrument makes me feel pretty good about the risk of any serious memory problems.
Now, let's contrast that with some code that has a seriously problematic "strong reference cycle":
Now this is a completely different picture, even though the process was the same "present and dismiss" process, repeated twice. This is now telling me that I had a 14mb growth between generations, and more notably, I can clearly see the problematic growth curve. What's remarkable is that while the Allocations tool is clearly catching a serious problem, the Leaks tool reports nothing.
Now, in practice, the real-world experience with the Allocations tool will probably rest somewhere between these two extremes. Your app may have its own caches or model objects that slowly take up memory, but if you're properly responding to memory warnings, you should be recovering that memory. Frankly, though, most well designed apps should not be generating memory warnings at all (usually accomplished by properly configuring caches, avoiding imageNamed where appropriate, moving to persistent storage for large or infrequently accessed data, etc.). The goal is to get to a point where the app stabilizes around some reasonable baseline memory allocation level, consistently returning back to that baseline.
But it's going to be impossible for us to advise you until you do some basic profiling of your app and diagnose the sorts of memory issues that you're having.
I recently ported a project over to ARC as I was having trouble with crashes and actually determining the cause, whether it was leaks or retain cycles etc., Now I have ported it over, I have not done massive testing to see whether it still crashes as I have not managed to get passed the activity monitor giving me the heeby jeebies when it shows my application doing This (activity monitor profiler)
whereas in allocations tools it looks something like
That real memory usage is not even the worst of it, at one point it shot up to around 90 odd MBs, I am unsure on how to proceed as I am not 100 percent sure what to do with the information given here, Except assume that I might be dong something, very wrong, And I have also run the leaks instrument, I have a few but they are minimal, they are all in bytes.
Does anyone have an explanation? or at the very least are able to clarify what I am possibly looking at? what's the difference between real memory usage and live bytes and overall bytes? Also these results were gotten doing exactly the same actions once and then showing you at the end of it.
I have been trying to reduce the real memory usage as pre ARC conversion I was having memory warnings and silent crashes frequently, I have not run into these again after converting, but I have not done any prolonged testing as I cannot conceive of even trying when the real memory usage looks like that. Which actually looks a lot higher than before ARC...Although the live bytes does look lower post ARC...Madness!
Something that confused me for a while is that ARC - wonderful as it is - does not necessarily avoid the need for #autoreleasepool.
https://developer.apple.com/library/mac/#documentation/Cocoa/Conceptual/MemoryMgmt/Articles/mmAutoreleasePools.html
I ran very large memory usage in an app until someone suggested:
#autoreleasepool {
// lots of allocating of objects returned from methods then discarded
} // and the closing brace of the autoreleasepool block causes their memory to be recovered here
Maybe that will help you.
A good explanation of the meaning of the various columns in the profiler is at Instruments ObjectAlloc: Explanation of Live Bytes & Overall Bytes
I am trying to work through some low memory conditions using instruments. I can watch memory consumption in the Physical Memory Free monitor drop down to a couple of MB, even though Allocations shows that All Allocations is about 3 MB and Overall Bytes is 34 MB.
I have started to experience crashing since I moved some operations to a separate thread with an NSOperationQueue. But I wasn't using instruments before the change. Nevertheless, I'm betting I did something that I can undo to stop the crashes.
By the way, it is much more stable without instruments or the debugger connected.
I have the leaks down to almost none (maybe a hundred bytes max before a crash).
When I look at Allocations, I only see very primitive objects. And the total memory reported by it is also very low. So I cant see how my app is causing these low memory warnings.
When I look at Heap Shots from the start up, I don't see more than about 3 MB there, between the baseline and the sum of all the heap growth values.
What should I be looking at to find where the problem is? Can I isolate it to one of my view controller instances, for example? Or to one of my other instances?
What I have done:
I powered the device off and back on, and this made a significant improvement. Instruments is not reporting a low memory warning. Also, I noticed that Physical Free Memory at start up was only about 7 MB before restarting, and its about 60 MB after restarting.
However, I am seeing a very regular (periodic) drop in Physical Free Memory, dropping from 43 MB to 6 MB (an then back up to 43 MB). I would like to knwo what it causing that. I don't have any timers running in this app. (I do have some performSelector:afterDelay:, but those aren't active during these tests.)
I am not using ARC.
The allocations and the leaks instruments only show what the objects actually take, but not what their underlaying non-object structures (the backing stores) are taking. For example, for UIImages it will show you have a few allocated bytes. This is because a UIImage object only takes those bytes, but the CGImageRef that actually contains the image data is not an object, and it is not taken into account in these instruments.
If you are not doing it already, try running the VM Tracker at the same time you run the allocations instrument. It will give you an idea of the type memory that is being allocated. For iOS the "Dirty Memory", shown by this instrument, is what normally triggers the memory warnings. Dirty memory is memory that cannot be automatically discarded by the VM system. If you see lots of CGImages, images might be your problem.
Another important concept is abandoned memory. This is memory that was allocated, it is still referenced somewhere (and as such not a leak), but not used. An example of this type of memory is a cache of some sort, which is not freeing up upon memory warning. A way to find this out is to use the heap shot analysis. Press the "Mark Heap" button of the allocations instrument, do some operation, return to the previous point in the app and press "Mark Heap" again. The second heap shot should show you what new objects have been allocated between those two moments, and might shed some light on the mystery. You could also repeat the operation simulating a memory warning to see if that behaviour changes.
Finally, I recommend you to read this article, which explains how all this works: http://liam.flookes.com/wp/2012/05/03/finding-ios-memory/.
The difference between physical memory from VM Tracker and allocated memory from "Allocations" is due to the major differences of how these instruments work:
Allocations traces what your app does by installing a tap in the functions that allocate memory (malloc, NSAllocateObject, ...). This method yields very precise information about each allocation, like position in code (stack), amount, time, type. The downside is that if you don't trace every function (like vm_allocate) that somehow allocates memory, you lose this information.
VM Tracker samples the state of the system's virtual memory in regular intervals. This is a much less precise method, as it just gives you an overall view of the current state. It operates at a low frequency (usually something like every three seconds) and you get no idea of how this state was reached.
A known culprit of invisible allocations is CoreGraphics: It uses a lot of memory when decompressing images, drawing bitmap contexts and the like. This memory is usually invisible in the Allocations instrument. So if your app handles a lot of images it is likely that you see a big difference between the amount of physical memory and the overall allocated size.
Spikes in physical memory might result from big images being decompressed, downsized and then only used in screen resolution in some view's or layer's contents. All this might happen automatically in UIKit without your code being involved.
I have the leaks down to almost none (maybe a hundred bytes max before a crash).
In my experience, also very small leaks are "dangerous" sign. In fact, I have never seen a leak larger than 4K, and leaks I usually see are a couple hundreds of bytes. Still, they usually "hide" behind themselves a much larger memory which is lost.
So, my first suggestion is: get rid of those leaks, even though they seem small and insignificant -- they are not.
I have started to experience crashing since I moved some operations to a separate thread with an NSOperationQueue.
Is there a chance that the operation you moved to the thread is the responsible for the pulsing peak? Could it be spawned more than once at a time?
As to the peaks, I see two ways you can go about them:
use the Time Profiler in Instruments and try to understand what code is executing while you see the peak rising;
selectively comment out portions of your code (I mean: entire parts of your app -- e.g., replace a "real" controller with a basic/empty UIViewController, etc) and see if you can identify the culprit this way.
I have never seen such a pulsating behaviour, so I assume it depends on your app or on your device. Have you tried with a different device? What happens in the simulator (do you see the peak)?
When I'm reading your text, I have the impression that you might have some hidden leaks. I could be wrong but, are you 100% sure that you have check all leaks?
I remember one particular project I was doing few month ago, I had the same kind of issue, and no leaks in Instruments. My memory kept growing up and I get memory warnings... I start to log on some important dealloc method. And I've seen that some objects, subviews (UIView) were "leaking". But they were not seen by Instruments because they were still attached to a main view.
Hope this was helpful.
In the Allocations Instrument make sure you have "Only Track Active Allocations" checked. See Image Below. I think this makes it easier to see what is actually happening.
Have you run Analyze on the project? If there's any analyze warnings, fix them first.
Are you using any CoreFoundation stuff? Some of the CF methods have ... strange ... interactions with the ObjC runtime and mem management (they shouldn't do, AFAICS, but I've seen some odd behaviour with the low-level image and AV manipulations where it seems like mem is being used outside the core app process - maybe the OS calls being used by Apple?)
... NB: there have also, in previous versions of iOS, been a few mem-leaks inside Apple's CF methods. IIRC the last of those was fixed in iOS 5.0.
(StackOVerflow's parser sucks: I typed "3" not "1") Are you doing something with a large number of / large-sized CALayer instances (or UIView's with CG* methods, e.g. a custom drawRect method in a UIView?)
... NB: I have seen the exact behaviour you describe caused by 2 and 3 above, either in the CF libraries, or in the Apple windowing system when it tries to work with image data that was originally generated inside CF libraries - or which found its way into CALayers.
It seems that Instruments DOES NOT CORRECTLY TRACK memory usage inside the CA / CG system; this area is a bit complex since Apple is shuffling back and forth between CPU and GPU ram, but it's disappointing that the mem usage seems to simply "disappear" when it clearly is still being used!
Final thought (4. -- but SO won't let me type that) - are you using the invisible RHS of Instruments?
Apple hardcoded Instruments to always disable itself everytime you run it (so you have to keep manually opening it). This is stupid, since some of the core information only exists in the RHS bar. But I've worked with several people who didn't even know it existed :)
I am Profiling an Application with Instruments. The profiling is done using Allocations Tool in two ways:
By choosing Directly the Allocations when I run the App for Profiling
By Choosing Leaks when I run the App for Profiling.
In both the cases , I had Allocations tool enabled for testing. But surprisingly , I had two different kind of Out put for Allocations at these instances.
Are they supposed to behave differently? or this is a problem with Instruments.
The time I Profile the with Leaks Tool:
In Allocations Graph:
1. I get lot of Peaks in Graph, The Live bytes and overall bytes are same.
2. I get the Black flags (I think it alarms about memory warning) after 1 minutes usage. Then after a set of flags appearing, my App Crashes. (This happens at times, even when directly run the App in Device)
The time I Profile the with Allocation Tool:
In Allocations Graph:
1. I do not get peaks often as it was in the above case. The Live bytes was always way less than Overall bytes.
2. I have used for over 20 minutes and never got Black flags.
One fact I came to know is that, when Live bytes and overall bytes are equal, the NSZombieEnabled could be enabled.
Have any of you ever encountered this problem.
UPDATE 1:
I faced another problem with first case. Whenever I profiled after a short duration (compared to profiling in second case), the app got lots of Black Flags and my App Crashed. (Due to Memory warning)
And when I tried the similar step by step use of Application my Application did not crash and got no flags.
Why this discrepancy?
In the first case, you are only tracking live allocations because the "Leaks" template configures the Allocations instrument that way. In the second, you are tracking both live and deallocated allocations. (As CocoaFu said).
Both are useful, but for slightly different reasons.
Only tracking live allocations (in combination with Heapshot Analysis, typically), is a great way to analyze permanent heap growth in your application. Once you know what is sticking around forever, you can figure out why and see if there are ways to optimize it away.
Tracking all allocations, alive and dead, is a very effective means of tracking allocation bandwidth. You can sort by overall bytes and start with the largest #. Have a look at all the points of allocation (click the little arrow next to the label in the Category of the selected row), and see where all the allocations are coming from.
For example, your graph shows that there are 1.27MB of 14 byte allocations -- 9218 allocations -- over that period of time. All have been free()d [good!], but that still represents a bunch of work to allocate, fill with data (presumably), and free each one of those. It may be a problem, maybe not.
(To put this in perspective, I used this technique to optimize an application. By merely focusing on reducing the # of transient -- short lived -- allocations, I was able to make the primary algorithms of the application 5x faster and reduce memory use by 85%. Turns out the app was copying strings many, many, times.)
Not sure why your app crashed as you described. Since it is a memory warning, you should see what is most frequently allocated.
Keep in mind that if you have zombie detection enabled, that takes a lot of additional memory.
Depending on the way Allocations is instantiated there are different options. Check the options by clicking the "i" symbol in the Allocations tile.
Yes, I find this annoying also.
My application keeps consuming more and more memory as seen in the Windows Task Manager and eventually crashes due to OutOfMemory. However when i check for leaks using MemoryValidator (from www.softwareverify.com) no leaks are detected. Why is this happening?
Just because there is a growing amount of memory usage doesn't mean it is necessarily 'leaking'. You could simply be accumulating a large number of live objects and/or very large ones (containing lots and lots of data).
If you can provide more information about what language(s) you are using and what the application is doing I can perhaps help out with some more specific information!
UPDATE AS PER COMMENTS
Well, you'll just want to make sure the garbage collection is happening correctly. I'd suggest the libgc library to help with that perhaps.
http://developers.sun.com/solaris/articles/libgc.html
The only other thing I could think of as being the cause of this is that you are maintaining references to the objects somewhere unintentionally so they are just piling up.