cancel
Showing results for 
Search instead for 
Did you mean: 
Create Post
Level 9

VMAN Snapshot Polling

Jump to solution

Our backup solutions uses snapshots in it's backup process, with this happening nightly we have a desire to send out a daily report to see if there are any snapshots that may have been left behind.

The issue has been that the report isn't accurate all the time.

Additionally, i've deleted Snapshots and after hours SolarWinds is still showing them.

I've been trying to research the polling interval for this data but can't seem to find any information that talks about it.

All of our polling seems to be working in VMAN and eventually the Snapshot data catches up (I'm not sure if it's X time later or if it's after the nightly database maintenance)

0 Kudos
1 Solution
Product Manager
Product Manager

elevate  wrote:

Our backup solutions uses snapshots in it's backup process, with this happening nightly we have a desire to send out a daily report to see if there are any snapshots that may have been left behind.

The issue has been that the report isn't accurate all the time.

Additionally, i've deleted Snapshots and after hours SolarWinds is still showing them.

I've been trying to research the polling interval for this data but can't seem to find any information that talks about it.

All of our polling seems to be working in VMAN and eventually the Snapshot data catches up (I'm not sure if it's X time later or if it's after the nightly database maintenance)

Happy holidays David,

We have configuration polling intervals and performance polling intervals as described here Virtualization Manager data collection and calculation of rollup information from VMware - SolarWind...  snapshot data would be collected along the configuration polling intervals, which can be indicative of the confusion you're describing, if you're looking for the reports to be reflecting changes almost immediately.

What intervals do you tend to run these reports? When do you want to know that these snapshots have been removed or remain behind?

The other thing I'd like to point out is that we actually have some updates to provide more real time alerting and reporting, and that is described in our beta forum VMAN Beta Now Available!  If you have a spare moment, please join us and tell us more about your use cases. I'd love to see if this feature can meet your needs on a faster cadence than the existing product features do.

View solution in original post

0 Kudos
9 Replies
Product Manager
Product Manager

elevate  wrote:

Our backup solutions uses snapshots in it's backup process, with this happening nightly we have a desire to send out a daily report to see if there are any snapshots that may have been left behind.

The issue has been that the report isn't accurate all the time.

Additionally, i've deleted Snapshots and after hours SolarWinds is still showing them.

I've been trying to research the polling interval for this data but can't seem to find any information that talks about it.

All of our polling seems to be working in VMAN and eventually the Snapshot data catches up (I'm not sure if it's X time later or if it's after the nightly database maintenance)

Happy holidays David,

We have configuration polling intervals and performance polling intervals as described here Virtualization Manager data collection and calculation of rollup information from VMware - SolarWind...  snapshot data would be collected along the configuration polling intervals, which can be indicative of the confusion you're describing, if you're looking for the reports to be reflecting changes almost immediately.

What intervals do you tend to run these reports? When do you want to know that these snapshots have been removed or remain behind?

The other thing I'd like to point out is that we actually have some updates to provide more real time alerting and reporting, and that is described in our beta forum VMAN Beta Now Available!  If you have a spare moment, please join us and tell us more about your use cases. I'd love to see if this feature can meet your needs on a faster cadence than the existing product features do.

View solution in original post

0 Kudos
Level 9

Thank you for your response Serena.

I took a look at the article you referenced to see if I can find where to configure the "configuration polling intervals" but couldn't seem to locate them in our environment. We're running 2018.4 and VMAN 8.3.1. Where can I configure this polling interval?

It definitely helps knowing that our intervals are set to 12 hours for this data and lowering this would help tremendously (although I'm not sure what resource implications that may have on our polling).

At this time we are only running the 1 report on a daily basis. If I can ensure the accuracy of this report that would be a good first step. At times, the recipients of this report are seeing snapshots that are created during the backup phase and have since been removed as the backup finished successfully, sending them looking for an issue that doesn't exist.

0 Kudos
Product Manager
Product Manager

elevate  wrote:

Thank you for your response Serena.

I took a look at the article you referenced to see if I can find where to configure the "configuration polling intervals" but couldn't seem to locate them in our environment. We're running 2018.4 and VMAN 8.3.1. Where can I configure this polling interval?

It definitely helps knowing that our intervals are set to 12 hours for this data and lowering this would help tremendously (although I'm not sure what resource implications that may have on our polling).

At this time we are only running the 1 report on a daily basis. If I can ensure the accuracy of this report that would be a good first step. At times, the recipients of this report are seeing snapshots that are created during the backup phase and have since been removed as the backup finished successfully, sending them looking for an issue that doesn't exist.

Hi again David,

Currently we don't have settings surfaced in the UI to allow you to tweak this because of the concern of the implications on your polling. This is a feature request that is referenced in part through this   which if you're interested, please give that a +1. That said, you can set the intervals in the database. If you change the polling intervals for an existing node, you'd want to take a look at Vim_PollingTasks table, but it's highly recommended that you do this with the help of a support representative, as I do not have insight into how your system (vCenter and VMAN) is performing right now. For instance, If you're running on the ragged edge of resources, this would not be recommended for you to try.

Now, to actually get to the problem you're solving, that's why I recommend that you take a look at the VMAN beta forum. In that beta, we're showcasing support for VMware events, which incidentally has support for snapshot events at a near real time rate. Near real time, is FAR better than trying to muck around with your polling intervals. I'd really love for you to set up a beta system and take a look to see if we could configure the beta to get you the report on the interval you really want.

out of curiosity, are you running the out of the box snapshot report or something more custom on your environment? If it's more custom, could you tell me what fields you've specified for the report?

Level 9

So I went to get you the fields we're looking at for our report and found that we are actually looking at locked VMDKs for this report.

pastedImage_0.png

In the report we are only showing the proxy servers ( name like pxyavp or pxyavb ) and we exclude any VMDKs that may match that same criteria as those we don't care about.

My apologies for the confusion / change of direction.

Do these locked VMDKs fall under the same 12 hour polling interval and if so is the Beta a good test for that logic?

Thanks,


David

0 Kudos
Product Manager
Product Manager

elevate  wrote:

So I went to get you the fields we're looking at for our report and found that we are actually looking at locked VMDKs for this report.

pastedImage_0.png

In the report we are only showing the proxy servers ( name like pxyavp or pxyavb ) and we exclude any VMDKs that may match that same criteria as those we don't care about.

My apologies for the confusion / change of direction.

Do these locked VMDKs fall under the same 12 hour polling interval and if so is the Beta a good test for that logic?

Thanks,


David

ah I see, that's interesting. Yes, this type of change would still fall under the 12 hour polling interval. The beta in that case probably won't be a good test for this logic, as it's not going to be tied to the events that I was considering for your use case. Have you considered changing the schedule of the report to coincide with the polling time of your configuration data? Do you have restrictions on when the report needs to be sent out?

Level 9

I'm trying to find out what time the configuration polling is being done at and found this, not sure if it's right or how I could change it to match when we want the report to run.

pastedImage_0.png

The above screenshot is from our Orion.VIM.PollingTasks table/view.

It seems as though some of our vcenters aren't even polling and other ones the 12-28 (Today was apparently done as 14:02 when it's only 13:12)

0 Kudos
Product Manager
Product Manager

elevate  wrote:

I'm trying to find out what time the configuration polling is being done at and found this, not sure if it's right or how I could change it to match when we want the report to run.

pastedImage_0.png

The above screenshot is from our Orion.VIM.PollingTasks table/view.

It seems as though some of our vcenters aren't even polling and other ones the 12-28 (Today was apparently done as 14:02 when it's only 13:12)

That is the correct table, but I'd recommend that you work with support to change those fields. If you Submit a Ticket | SolarWinds Customer Portal then you'll be able to work with some of our support reps to set that appropriately.

As far as the timestamp, that one is in UTC (most tables are, but not everything unfortunately).

- Marc Netterfield, Github
Level 9

we have a few different vcenters, I imagine their polling intervals are staggered a bit which could make it hard for me to link everything up.

What is the best way for me to go about finding the configuration polling time for each vcenter?

This report is scheduled to go out after all of our backups have finished. We've scheduled the report to run at 10:30 as that should have been enough time for any stragglers to finish up. I suppose, if we could get all vcenters to poll at 10:00am and then send out the report 30min later that would make it reliable?

0 Kudos