eG Monitoring
 

Measures reported by RedisCpuTest

CPU contentions on the Redis server can significantly degrade the performance of the server. To avoid this, administrators must track the CPU usage of the server over time, proactively detect potential CPU contentions, and accurately isolate what is causing excessive CPU usage on the server - main/background system processes? main/background user processes? This can be achieved using the RedisCpuTest.

This test monitors the CPU usage of the Redis server and alerts administrators to erratic usage patterns on the server and what is causing them. This way, the test helps administrators isolate and pre-empt CPU contentions on the server, and thus improve server performance.

Outputs of the test : One set of results for the target Redis server

The measures made by this test are as follows:

Measurement Description Measurement Unit Interpretation
used_cpu Indicates the percentage of CPU resources used by the target server Percent A value close to 100% is a cause for concern, as it implies that the server is using CPU resources excessively. If the situation persists, the server may soon not have enough processing power to process requests to it
used_cpu_sys Indicates the duration for which system processes utilized CPU. Seconds This is the sum of system CPU consumed by main threads and background threads of the server process.

By comparing the value of this measure with that of the used_cpu_user measure, you can instantly determine whether system processes are CPU-hungry or user processes.
used_cpu_sys_children Indicates the period of time for which background system processes hogged the CPU. Seconds If the value of the used_cpu_sys measure is abnormally high, you may want to determine what type of system processes are CPU-intensive -background processes? main processes?

While the value of this measure will indicate how much CPU was used by background system processes, you can quickly compute the CPU usage of main system processes using this formula:

used_cpu_sys (-) used_cpu_sys_children

Then, compare the value of this measure with the result of the above computation to figure out the rootcause of the excessive CPU usage by system processes.
used_cpu_user Indicates the duration for which user processes hogged CPU resources. Seconds This is the sum of user CPU consumed by main threads and background threads of the server process.

By comparing the value of this measure with that of the used_cpu_sys measure, you can instantly determine whether system processes are CPU-hungry or user processes
used_cpu_user_children Indicates the period of time for which background user processes hogged the CPU. Seconds If the value of the used_cpu_user measure is abnormally high, you may want to determine what type of users rocesses are CPU-intensive - background processes? main processes?

While the value of this measure will indicate how much CPU was used by background user processes, you can quickly compute the CPU usage of main user processes using this formula:

used_cpu_user (-) used_cpu_user_children

Then, compare the value of this measure with the result of the above computation to figure out the rootcause of the excessive CPU usage by user processes.