Developping time and time waste
Hello, I don't know if it's a real question (because I don't know if there is real answers...) but... I noticed that now, working on my project suite means waiting about half of the time for ...
Forum Discussion
Have you tried copying the log directory to your local and attaching to TC from there? That would at least rule in / out network latency as the cause of your issue
Logs are shared by numerous colleagues...
I must put it on a shared place.
Did you check to make sure that your virtual machines meet or exceed the minium requirements for Test Complete?
https://smartbear.com/product/testcomplete/specifications/
Also consider memory leaks, I dont know what you are testing but I know for our web testing we need to restart IE every 30 minutes or so during a test run as either IE and or Test Execute ties up so many resources that the application becomes unresponsive. You might want to check how much resources are in use when Test Execute 'freezes'.
chrisb has a good point with memory leaks.
I'm currently testing a Delphi application. Certain objects are transient during a run, cannot be reliably mapped, and have to found during a run.
One such item is a popout menu that is generated from a TreeView node, which can be several levels deep. Each level has to be searched for at runtime as it is spawned.
I noticed it taking longer and longer to find them.
When I inspected the object model, there were thousands of ADODB connection objects floating around in the background that the application was not disposing of properly. It had probably been doing this for years. They were eventually killed when the module was closed (the module is launched from out primary application).
But no-one had ever noticed this. On a larger scale (such as if it had been happening in the main application rather than the smaller Delphi executable it launches) this could have led to far worse memory consumption and performance. Have seem similar things with web based applications as well. Worst one was capable of locking up my machine (i7 with 8gb RAM at the time) with 40 minutes of constant running as it was so bad.
So yeah, definitely something you should consider looking into. Especially on a VM where problems like this can be magnified if they are not allocated enough resource both in themselves and their hosting environment.
Colin_McCrae Thats interesting. I hadn't thought of the difference between name mapping and finding objects on the fly. I did away with name mapping and find all my objects on the fly as the test is running. I'd assume finding them on the fly is way more resource intensive. Perhaps this is something SmartBear doesnt even test too much as they seem to encourage users to use namemapping.
Yes but that does not stop you (or your colleagues) COPYING them to your local
TheGhost I don't understand your suggestion. Perhaps you are sughesting that running on the VM might be the issue and they can try on a desktop too? I'd still suggest adding a script to restart the browser peridocially. If that works then you know you need to start look at resouce consumption as your application runs over time.
Sorry must have hit the wrong reply button - reply was to the post above. Not implying your suggestion is not valid.
When logs get really big accessing them over a network share can slow browsing the log down
