Live Optics Viewer - Environment Summary for Top Five Servers - what do all the #s mean?
I cannot find clear information on what the information reported under the "Top Five Servers" section on the Environment Tab of a project means. The reported information is ambiguous - are the %s averages? (seems so since they differ from peak), is the nominal number just the % * the available or is it something else?
Can someone please explain what the %s & then the nominal number that shows up on a hover-over mean - for each of the 5 reported categories?
The hover-over number represents the data point which created the bar in the graph. For example. If you have Top 5 IOPS selected, and hover over one of the bars to reflect 1050, that servers peak IOPS was 1050. If you click on it, it will take you to that server in the Performance Tab.
The percentage should add up to 100% and reflects a bug. I will have engineering take a look. Thank you for pointing it out.0
Here is an explanation for each. The first 2 do not add to 100 %, the other 3 will add to 100%.
Percentage = Percentage of the Total Memory for the project that the peak of this server consumed. e.g. Total project of 144 GB, Peak for Server of 35GB = 25%
Hoverover value = Peak memory usage.
Percentage = Similar to Memory. Percentage of the Project Net Clock Speed that the peak consumed e.g. Total project net clock speed of 41.2GHz, Peak for a server of 1.8 = 4%
Hoverover value = Peak CPU in GHz
Percentage = The Peak Aggregate Network Throughput of this server as a percentage of the sum of the Peak Aggregate Network Throughputs for all severs in the project. E.g. Server 1 Peak aggregate of 40.36 megabits/s, Server 2 of 22.75. Server 1 scores 64% for Network throughput.
Hoverover value = The Peak Aggregate Network Throughput (sum of recieved + sent) of this server
Percentage = The sum of average read + average write of this server as a percentage of the sum of all the average read & write for all servers.
Hoverover value = Average Read + Average write in ms
Percentage = Similar to above. The peak IOPS of this server as a percentage of the sum of all the servers peak iops.
Hoverover value = Peak IOPS0
Thank you for the replies. I am looking for average values across many of the parameters such as CPU, MEM, etc. It seems that LOVA is only capturing and basing the model on peak. Is there a way to get more statistics than peak, such as average? If so, please instruct where to find that.0
Hello - pinging again on this as it's critical for my project to understand:
It seems that LOVA is only capturing and basing the model on peak. Is there a way to get more statistics than peak, such as average? If so, please instruct where to find that.0
Sorry, Steven. The averages are an active Feature Request, but not currently available.0
For individual server Avg Latency, Avg IO size and Avg Queue depth can be found under the performance tab, go to that individual server, expand more information, info on the right hand side. IOPS, throughput, CPU and memory do not have avg values at this time.0
Thanks, there are some average values in that area, but none I can find around CPU, memory, and networking that I can find.0
Please sign in to leave a comment.