This question is locked. New answers and comments are not allowed.
Anyone else having the issue where when you take a snapshot, the application you are profiling becomes unresponsive after the snapshot is complete?
We are trying to investigate a memory leak and cannot compare snapshots because we can never get more than a single snapshot. After the first the application cannot be interacted with.
We are trying to investigate a memory leak and cannot compare snapshots because we can never get more than a single snapshot. After the first the application cannot be interacted with.
6 Answers, 1 is accepted
0
Hello Nathan,
Thank you for contacting us. I am sorry for the problem you are experiencing with JustTrace. I noticed that you are not using the latest version of JustTrace. After the Q3 SP1 was released we fixed an issue which in some cases crashed the profiled process when getting a memory snapshot. I think that you may be facing this bug.
I suggest that you upgrade to the latest internal build and retry profiling your application with it. To upgrade please go to the "Help & Support" page and press the "Check for Updates" button.
All the best,
Martin
the Telerik team
Thank you for contacting us. I am sorry for the problem you are experiencing with JustTrace. I noticed that you are not using the latest version of JustTrace. After the Q3 SP1 was released we fixed an issue which in some cases crashed the profiled process when getting a memory snapshot. I think that you may be facing this bug.
I suggest that you upgrade to the latest internal build and retry profiling your application with it. To upgrade please go to the "Help & Support" page and press the "Check for Updates" button.
All the best,
Martin
the Telerik team
0
Ristogod
Top achievements
Rank 2
Rank 2
answered on 13 Feb 2013, 04:39 PM
Thanks. I updated to the latest internal build.
So far I am unable to determine if the issue is fixed as the performance overall has reduced dramatically. A running process prior would take around 1-2 minutes with JustTrace profiling it. Now with the latest version, it is taking 1-2 hours.
So I'm not sure how useful it's going to be as it will most likely take me days to figure out what's going on.
0
Ristogod
Top achievements
Rank 2
Rank 2
answered on 13 Feb 2013, 07:00 PM
I said 1-2 hours prior. But in reality that was just a guess. I went to lunch and came back and it's still running. So it's likely more along the lines of 3-4 hours or more for something that runs in 1-2 minutes when JustTrace isn't profiling it.
So this version is just as broke as the other version. I'll have to look for alternative methods to help isolate where my memory leak is coming from. I really hope you guys can sort this out though and make your product useful again.
So this version is just as broke as the other version. I'll have to look for alternative methods to help isolate where my memory leak is coming from. I really hope you guys can sort this out though and make your product useful again.
0
Hi Nathan,
We routinely test JustTrace against different workloads and profile targets and this behavior is unexpected. We need more data to resolve the issue. Is it possible to send us the the project (the binaries should be enough) so we can test against it? Another options is to send us a memory dump of the profiled process while it is working. We may also arrange an on-line meeting to debug the problem on your machine, if this is convenient for you.
In all cases, I suggest you to turn off the "Collect Call Stacks' option to speeds up the memory profiler while collecting enough information to solve large classes of memory leak. Also, running on a physical machine (and not inside a virtual one) also speeds up the profiling.
Kind regards,
Todor
the Telerik team
We routinely test JustTrace against different workloads and profile targets and this behavior is unexpected. We need more data to resolve the issue. Is it possible to send us the the project (the binaries should be enough) so we can test against it? Another options is to send us a memory dump of the profiled process while it is working. We may also arrange an on-line meeting to debug the problem on your machine, if this is convenient for you.
In all cases, I suggest you to turn off the "Collect Call Stacks' option to speeds up the memory profiler while collecting enough information to solve large classes of memory leak. Also, running on a physical machine (and not inside a virtual one) also speeds up the profiling.
Kind regards,
Todor
the Telerik team
0
Ristogod
Top achievements
Rank 2
Rank 2
answered on 01 Mar 2013, 02:51 PM
I'm sorry I wasn't able to take the time to work with you in resolving any prior issues with JustTrace. I simply didn't have the time to deal with it directly.
However, I recently updated to your latest release and it went from being a 100% unusable product, to a completely usable product. I can now profile my application and it barely takes any more time to run while profiling than it does to run with JustTrace off. Not only that, but the snapshots button is now working correctly and it no longer locks up anything.
So I commend your efforts and thank you for resolving the issues.
However, I recently updated to your latest release and it went from being a 100% unusable product, to a completely usable product. I can now profile my application and it barely takes any more time to run while profiling than it does to run with JustTrace off. Not only that, but the snapshots button is now working correctly and it no longer locks up anything.
So I commend your efforts and thank you for resolving the issues.
0