VM Uptime very low even of the VM is accessible

[sonic23]

Hi,

I have noticed this issue when we upgraded from VeeamOne v7 to 9 that our VM Uptime report is odd. In the report I have a VM with 7 restarts plus 7 alarms but the uptime is 100% (says 0 downtime) and I have a VM with 0 restart and 6 alarms and gota 52.72% uptime, weird that the report says it down for 14 days when this server is always up and running. Also if I want some VMs not to be included in the report how can I isolate them?

Thanks.

[Shestakov]

Hi Jesus.

What was the interval parameter in the report and when did you make an upgrade?
Are you in VMware or Hyper-V infrastructure?

To isolate particular VMs, you can create Business View groups and build the report for chosen groups only.

[sonic23]

Hi Nikita,

The interval parameter is "Past Month", we're on VMware environment. I think we upgraded a month ago. I also read in some forum that when VeeamOne is down it will affect the VM Uptime as well?

I'll check the Business View Group and will let you know.

Thanks.

[Shestakov]

What if you make the report only on interval you use v9, does it look normal?
There was uptime measurement change since v9, now Veeam ONE uses its own metric.
Thanks!

[sonic23]

I tried changing the intervals but it doesn't make any difference. I was able to find the days that the uptime was bad but the thing is that the infrastructure was not really down.

If this keeps happening is it possible to just export the data and configuration of my VeeamOne and just reinstall it fresh?

[Shestakov]

Yes, it is possible, but I would ask you to contact Veeam support instead to find out the issue. It may be just this report since it really changed its behavior since v7 and re-install will not help. It may be something else, so please contact support team if you are disturbed.
As for the report it used to count uptime detecting events and performance, now there is a calculated metric used for the report and events just shown for troubleshooting.

[sonic23]

Hi Nikita,

I installed a new VeeamOne server to make sure the issue has nothing to do with my upgrade. I tried checking the VMUptime again, first it shows 100%....then a few minutes it went down to 98.5%...went down again to 95.3%....after an hour went back to 100%...now back to 99.3%. I assume its connected to the time of syncing data, but what I dont get is why in the report it shows as "Downtime" when there was no downtime happened on the infrastructure. The downtime value is the same for every VM. Do I need to tweak something as it's a report I need to submit every week.

Thanks.

[Shestakov]

Hi Jesus,

I tried checking the VMUptime again, first it shows 100%....then a few minutes it went down to 98.5%...went down again to 95.3%....after an hour went back to 100%...now back to 99.3%.

Were you getting these results from the report? That`s strange if it was 95.3% it should not become 100% later that day.

Could you contact the support team as your case is really interesting for us. Thanks

[Vitaliy S.]

Another quick test I would suggest to do: can you open a Monitor Client and check if you have any performance data for the VMs in question? Please select the same time interval as you did in the report.

[sonic23]

Hi Nikita,

I agree with you. In what parameter does exactly VeeamOne getting the "Downtime" data? I assume that if the VMs are accessible or up, VeeamOne shouldn't be declaring any downtime. I already have an open Ticket for this issue: 01928537.

Vitaliy,

I am checking the VMUptime Report for the current day, I used the day before as the Start Date and today as the End Date (as VeeamOne dont want the start and End date the same). That's where the "Uptime" changes as the "Downtime" parameter keeps on changing. There are VMs with "0" restart and "0" alarms but still gets 99.25% uptime..weird also as the "downtime for all VM is the same.

[Shestakov]

It depends on the hypervisor. We use vSphere we mostly rely on VM Heartbeat. The infelicity may be caused by Veeam ONE services interruptions(most likely in your case), performance data aggregations in the beginning and end of the observed interval.
Thanks!

[Vitaliy S.]

Thanks for the case ID, I will ask our dev team to take a look.