Comments
Sort by recent activity
Hey, don't worry about the word wrap - just found it on the output pane's context menu. Be worth adding an option to the toolbar as well though. / comments
Hey, don't worry about the word wrap - just found it on the output pane's context menu. Be worth adding an option to the toolbar as well though.
Cheers Eddie - appreciate it. / comments
Cheers Eddie - appreciate it.
No stress - thanks Eddie. / comments
No stress - thanks Eddie.
Yes, although I wouldn't guarantee this as a universal answer. In our case in was due to a virtual NIC misconfiguration in one of our VMs, which was running on VMWare VSphere. To fix the problem we needed to disable receive side coalescing on the guest NIC. To do this you can run the following from an elevated Powershell prompt: Disable-NetAdapterRsc * As I say, this fixed the issue for all, and all our troublesome queries started running in around 1ms, but I wouldn't guarantee it will work for everyone. HTH. Bart / comments
Yes, although I wouldn't guarantee this as a universal answer.In our case in was due to a virtual NIC misconfiguration in one of our VMs, which was running on VMWare VSphere.To fix the problem we n...
Thanks Jessica! / comments
Thanks Jessica!
Hi Paul,
Great, thanks. 4 GB should be fine, although .NET processes tend to start playing up at 1.7-1.8 GB on a 32-bit operating system; on a 64-bit OS this shouldn't be an issue at all.
If the profiler isn't using that much memory then it'll be because down to fragmentation occurring; that's the issue I talked about with the need for the 20 MB contiguous block for the memory mapped file. The thing is that generally fragmentation doesn't become a big issue until the process has swallowed more than 1 GB, so that would still be somewhat odd--there is a very remote chance that it could be an issue at any time though, so maybe it's just that.
If on the other hand you're seeing extremely high profiler memory usage then it could be one of the following:
- Extremely complex/very deep call stacks; it's on by default anyway, but if needs be I'd recommend switching on the "Simplify very complex stack traces to save memory".
- We have a memory leak, although I'd be surprised if the memory usage could build up enough to cause a problem in only ten minutes. Even with complex applications it's extremely rare for ANTS Profiler 4 to use more than 500-600MB on my system, and yours easily has enough headroom to cope with that.
Hope some of that is helpful.
Thanks,
Bart / comments
Hi Paul,
Great, thanks. 4 GB should be fine, although .NET processes tend to start playing up at 1.7-1.8 GB on a 32-bit operating system; on a 64-bit OS this shouldn't be an issue at all.
If the pr...
Hi Paul,
First off thanks for sending through the log files; they basically confirmed what Andrew suggested to me after I read your post earlier.
It looks like either ANTS Profiler is running out of memory, or is unable to allocate a contiguous 20 MB block for a memory mapped file. We have seen somewhat similar problems before in testing, and so normally I'd suggest the latter, but it's interesting that the OutOfMemoryExceptions are occurring in the timeline; that could be a side-effect though. Do you happen to have any idea how much memory ANTS Profiler was using when this happened? I'm slightly concerned that in the problem file what we're seeing is only about 15 minutes of usage, which isn't a great deal. Also (apologies if I've asked before), how much memory do you have installed?
I will say that restarting the profiler and kicking off your profiling session again should fix the problem, although that's clearly not ideal because you're going to lose the data from the session you've been running.
Thanks,
Bart / comments
Hi Paul,
First off thanks for sending through the log files; they basically confirmed what Andrew suggested to me after I read your post earlier.
It looks like either ANTS Profiler is running out o...
Hi Paul,
Thanks for this; it's not something we've seen before so I'm not sure why it's happening but we'll definitely take a look into it. Any chance you could send in your log files to bartDOTreadATredHYPHENgateDOTcom please?
Thanks,
Bart / comments
Hi Paul,
Thanks for this; it's not something we've seen before so I'm not sure why it's happening but we'll definitely take a look into it. Any chance you could send in your log files to bartDOTrea...
I think this is an experimental thing, however, you might try View > Options, and then switch "Optimization" to ".NET 3.5" to see if that makes any difference.
Hope that helps.
Thanks,
Bart / comments
I think this is an experimental thing, however, you might try View > Options, and then switch "Optimization" to ".NET 3.5" to see if that makes any difference.
Hope that helps.
Thanks,
Bart
Hi Eddie,
Thanks very much - appreciate it.
Couple of other things:
- It would be great if double-clicking a node opened up the source code,
- Similarly to the context menu Select > Connected Objects option, a Show > Connected Objects would be really helpful (or, better yet Show > Objects This Depends On, and Show > Objects That Depend On This).
(I know, give 'em an inch...)
Thanks! / comments
Hi Eddie,
Thanks very much - appreciate it.
Couple of other things:
- It would be great if double-clicking a node opened up the source code,
- Similarly to the context menu Select > Connected Objec...