I want to profile my app but I'm not sure what every thread do. Of course I can suppose about the most "meaningful" threads role but some threads responsibility are still unclear for me. E.g. what is "crap"-thread and "Studio:socket" threads? Also what threads are deserve attention while profiling? By the way it's really strange for me that number of threads are not the same for emulator and physical devise:
Emulator, Run:
Physical device, Run
Where can I find descriptions of these threads? I had tried to Google it but without success.
The Studio: threads are threads started by Android Studio's JVMTI agent, useful for capturing performance-related app data and sending them to the Android Studio profiler. If you don't want to see them or want to minimize their performance impact on your app, you can build your app with the profileable tag. This way the JVMTI agent won't be attached but you will lose some profiler functionality.
As for the crap.coldshower thread, it's a truncated string because Linux thread names are capped at 15 characters. The full name is likely some package name com.xxxcrap.coldshower, usually the main thread of an app.
Related
On the Compute Engine VM in us-west-1b, I run 16 vCPUs near 99% usage. After a few hours, the VM automatically crashes. This is not a one-time incident, and I have to manually restart the VM.
There are a few instances of CPU usage suddenly dropping to around 30%, then bouncing back to 99%.
There are no logs for the VM at the time of the crash. Is there any other way to get the error logs?
How do I prevent VMs from crashing?
CPU usage graph
This could be your process manager saying that your processes are out of resources. You might wanna look into Kernel tuning where you can increase the limits on the number of active processes on your VM/OS and their resources. Or you can try using a bigger machine with more physical resources. In short, your machine is falling short on resources and hence in order to keep the OS up, process manager shuts down the processes. SSH is one of those processes. Once you reset the machine, all comes back to normal.
How process manager/kernel decides to quit a process varies in many ways. It could simply be that a process has consistently stayed up for way long time to consume too many resources. Also, one thing to note is that OS images that you use to create a VM on GCP is custom hardened by Google to make sure that they can limit malicious capabilities of processes running on such machines.
One of the best ways to tackle this is:
increase the resources of your VM
then go back to code and find out if there's something that is leaking in the process or memory
if all fails, then you might wanna do some kernel tuning to make sure your processes have higer priority than other system process. Though this is a bad idea since you could end up creating a zombie VM.
I'm developing Windows Metro App and in my App I need to download some information(about 60Mb, every time) from server in background. Download should occur regularly, for example every 8 hours. So I tried to use Background Task, but there are some CPU and network quotas(https://msdn.microsoft.com/en-us/library/windows/apps/xaml/hh977051.aspx), and I can't do this. Could somebody help me with advice in this problem?
Instead of attempting to do the entire transfer in the background task itself, have the background task start a background transfer, which runs independent of the task and independent of the app as well. See the topic, Transferring Data in the Background, https://msdn.microsoft.com/en-us/library/windows/apps/hh452979.aspx. You can run the background task periodically according to the schedule you need, and it will easily stay under CPU and networking quotas because the background transfer doesn't count against that.
I am developing an application that calls a web service, which deletes information from a database (the web service was developed by a third party vendor). On the first run approximately 100,000 records are deleted.
I have tested the routine a few times and this appears in Visual Studio occasionally:
"The CLR has been unable to transition from COM context 0x22c4f60 to COM context 0x22c51b0 for 60 seconds.
The thread that owns the destination context/apartment is most likely either doing a non pumping wait or processing a very long running operation without pumping Windows messages.
This situation generally has a negative performance impact and may even lead to the application becoming non responsive or memory usage accumulating continually over time.
To avoid this problem, all single threaded apartment (STA) threads should use pumping wait primitives (such as CoWaitForMultipleHandles) and routinely pump messages during long running operations."
I assume that the web service is taking more than sixty seconds to pass control back to the .NET Forms app. Please see the following quote from the message: "To avoid this problem, all single threaded apartment (STA) threads should use pumping wait primitives (such as CoWaitForMultipleHandles) and routinely pump messages during long running operations". As this is a Windows Forms app, does this mean that I do not need to do anything to allow for this?
Sometimes this issue may also occur due to wrong server name, like if you had included SERVER/SQLEXPRESS instead of SERVER\SQLEXPRESS then also this error displays, which was my case.
Check and be sure if you are using a reader that you do not included try/catches. You should make and effort to try and resolve your solution issue better than that. A try/catch will cause a time out especially in a while loop.
We have a few applications which are running in Windows 2K, 2008 servers. They are written in java.
These applications needs to do many automation tasks. We are having difficulty to monitor these applications. Sometime due to XYZ reasons application either hangs or fail to perform desired job. We only come to know about this after a few days when some one reports that desired function hasn't been executed.
To come out of this issue, we added emails for each imp exceptions but then developer needs to spend time to check those 1000 emails everyday. Which is again not feasible & efficient solution.
Now we are looking for a alert, alarms, notification display & monitoring system. We need to have a remote application which can receive alarms from these java applications & then based on certain information/Condition/Configuration, remote application can display some red, orange, green text on the screen. Based on red text, users can be visually see that there is an issue in the system. If required users can be notified that there is a serious issue in the application.
Please help us to identify any existing mechanism, tool, package to achieve this goal. Any suggestion would be highly appreciated.
Thanks
There are myriad ways to achieve this, but all of them will require some effort. Which way to proceed depends on your needs and abilities. A couple of options occur to me:
Have your processes log their exceptions to a Syslog daemon, running on some central server. Then you could have an admin read through the log file for serious problems, but there are many ways to post-process syslog messages, a web search on it might give some more hints.
Is there any way, when logged into the server, to observe whether or not the process is running properly or not? You could install something like Nagios on a sever, and write a plugin that monitors your particular process on all the servers. The plugin can basically be a shell script that checks "ps", or a log file, or whatever you want.
If you are in an IT department, your organization might already have some system like this (NMS).
I'm not sure why this question is tagged "snmp", but it's technically possible to install an SNMP agent on each server, and have them send traps on certain conditions. I do think it would be slightly overkill because you would also have to get a good SNMP manager to receive the traps and alert a sysadmin.
I would go for a combination of the check_logfiles plugin to parse log exceptions and raise alerts, and check_jmx/jmxquery to check metrics inside the JVM such as heap usage and thread count.
check_logfiles
check_jmx
I need to modify the number of threads available in my Weblogic 10.0 MP2 environment for some perf benchmarking but I cannot seem to be able to find where exactly that option lies.
Can anyone share this info please? thank you.
Weblogic 10 does not use execute thread queues like in previous versions (i.e. Weblogic 8.1 and older)
This concept is now replaced with Work Managers.
These are self-tuned, i.e. WLS will auto-tune the number of threads every 2 seconds based on how it sees the need to increase threads for the application load.
You can confirm this from the console, it will show the increasing number of execute threads as the load increases.
You can use the work manager and constraints to make sure your applications get certain criteria met.
Such as certain web apps or EJBs can get a higher share of threads and so on.
For a quick read see http://www.oracle.com/technetwork/articles/entarch/workload-management-088692.html
and
http://m-button.blogspot.com/2009/02/tuning-default-workmanager-on-weblogic.html
Secondly, are you running in dev mode or production mode.
If dev mode, you can try this cmd line parameter
-Dweblogic.threadpool.MinPoolSize=100
but I am not sure if it will work, so it's better to leave it to Work Managers