![]() As your current ASP is LinuxFree, Would recommend you to split up your apps and move them to another App Service plan or scale up App Service plan.So for each App running on an App Service Plan, it could be possible for up to four (4) containers to be running depending on the features which are enabled.If Managed Identities is used, a separate container will be added.If CORS or Authentication features are used by the App, another middleware container will be added to support the feature.Each App will have a corresponding Kudu site which is used for deployment and troubleshooting.The App Service and underlying features each run within their own containers.īelow is additional information regarding this topic. You may want to know that the App Service Linux and Web App for Containers uses Docker. As you mentioned that the issue did not exist previously so it might be helpful to review any recent changes to the code or to the Azure environment to determine if there are any potential causes for this issue.įurther looking at your app, there is too many actively running containers (>=10) detected leading to high CPU usage.This might involve reducing the number of data points being visualized, using more efficient algorithms, or caching the results to reduce the number of calculations required. To prevent this function from using excessive CPU resources, you could try optimizing the code or reducing the complexity of the visualization process.AFAIK "vizplugins" is a function that's used to visualize data or to generate graphs or charts."vizplugins" is not a standard Python function, so it's possible that it's a custom function that's been added to your application.after_cpu_total = user + nice + system + idle + iowait + irq + softirqĥ.If I have understood right it seems like the continuous CPU usage issue in your Azure web app is related to the "vizplugins" function.prev_cpu_total = user + nice + system + idle + iowait + irq + softirq.Calculating total cpu usage for beginning and end values: System:Processes executing in kernel modeĤ.Nice:Niced processes executing in user mode.User:Normal processes executing in user mode.Reading below values from the first line of /proc/stat file: Note: Since shell commands(in this case hash calculation command) are executed as child process in the shell script, we have to consider to calculate the total cpu usage of the process.ģ. after_proc_cpu_total = utime + stime + cutime + cstime.prev_proc_cpu_total = utime + stime + cutime + cstime.Calculating process’ total cpu usage for beginning and end values: Cstime (#17): Amount of time that this process’s waited-for children have been scheduled in kernel mode, measured in clock ticksĢ.Cutime (#16): Amount of time that this process’s waited-for children have been scheduled in user mode, measured in clock ticks.Stime (#15): Amount of time that this process has been scheduled in kernel mode, measured in clock ticks.Utime (#14): Amount of time that this process has been scheduled in user mode, measured in clock ticks.Reading below values from /proc/pid/stat file: (# indicates the column number in the stat file) Calculating CPU usage of the process for the specific time interval Calculation with proc filesġ. Reading /proc/stat and /proc/pid/stat files at the beginning and end of the hash calculation(pid value is the process id of the shell script which executes checksum N times)Ĥ. Creating shell script that runs given hash algorithm N timesģ. txt files according to given size and countĢ. Creating random(for obtaining different checksum results) dummy test. ![]()
0 Comments
Leave a Reply. |