Hi,
we are running a fresh install of vcops now for more than 30 days.
But i have a lot of VM´s and host where on the Operation -- Details page the status is "not calculated yet"
Also i have an VM where i wan´t to analyse the Anomalies. For me, anomalies is something that i could not really get.
Hopefully someone could give me some information about how to analyse a vm with a high anomolies value.
Attached, one vm that i wan´t to analyse deeper. Because it is an Database machine.
I appreciate any helpful comments.
Thanks
Frank
1)
Anomalies badge going "up" (above 0) basically indicate that there's something anomalous / out of the ordinary going on with your machine.
I think that's pretty obvious in the first snapshot - all was "calm" and at certain point the CPU usage spiked quite a lot. From what I see it used to be somewhere between 0 (idle) and ~15%,
and now it's 24%.
The items on the right are indicators what else has changed at the same time. I see spikes in network thruput, disk I/O, CPU usage etc - so that VM basically started doing something out of the blue,
most likely after it got some information over the network.
That's what I see on that first snapshot
2)
As to the "not calculated yet" - Is that only for the Anomalies badge, or for other badges as well? I think for Workload you should see the "Normal X-Y", cause there are DT indicators on the snapshots
(that's the small-blue-things above the each of the 4 green stripes in the "workload" section.It tells you the range where each value runs over time)
What it means is that although there's data for 1 month, the system haven't been able to find a valid fit/trend within the available data, in order to tell you what should be normal
Hi,
thanks!
Yes, you are right, the "not calculated yet" is only for the Anomalies badge. all other badges are calculated and have a value.
For me, the anomalies are one of the "not so easy to understand" thinks.
Is it the blue line showing in the anomalies badge is that what the vm or host is using at a specific time right?
The values that are displayed in the anomalies badge, how long did they be preserved?
And when did the value for the anomalies badge change with the knowledge about the different values of datastore, disk, cpu etc.?
And what is if there is no blue line? Is it then using 100%? That could not be right?
Do you now how long does it take until the "normal" value will be calculated?
Appreciate your help
Frank
man, you have a thousand questions
- Is it the blue line showing in the anomalies badge is that what the vm or host is using at a specific time right? -> not sure that I understand the question
- The values that are displayed in the anomalies badge, how long did they be preserved? -> you mean the symptoms on the right pane?
They will be shown while they're still considered active. So they'll most likely go away soon
- And when did the value for the anomalies badge change with the knowledge about the different values of datastore, disk, cpu etc.? -> because for them there is also 'dynamic range' being
calculated. And they went off. And because they "went off", the anomalies badge went > 0. Consider it as a summary, that for this VM something anomalous (out of the ordinary) is happening.
See snapshot
- And what is if there is no blue line? Is it then using 100%? That could not be right? -> look at my picture. The blue thing is not a line, it's a range - from A to B.
It is telling you that for the past month, for most of the time, that metric moved only inside that range. e.g. in the case of CPU, was between 0 and 15% usage.
And right now it's outside that range, which is an anomaly (something different from the normal behaviour)
Hi,
yes you are right, i have more then thousand questions .
For the blue line, please look at my screenshot. Sorry for that, i have to show it in my question before.
That is what is currently used from 100 % right?
Frank
The 100% in your screen shot is the percentage of objects for which a particular metric is tracked, and that metric is violating its normal values. So, since you're looking at symptoms for only a single resource, everything will be 100%.