Is there any way we can create dynamic groups and group names .
Computers not assigned to any groups need to grouped in 50’s and assigned to a new group (With dynamic name).
There is no way in the Console that I know of to do what you are asking for.
You might be able to use the REST-API to automate it though.
Can you provide me some documentation to REST API .
Also, if you simply Google “BigFix Rest API” there are several good links.
What is the end goal? Why do you need dynamic groups in blocks of 50 machines?
You can use relevance to approximate this automatically without the overhead of creating a group by using computer id mod ## = 0 for group 1 and so on.
In general, it wouldn’t be a good idea to make tons of groups, especially manual groups.
Also, with the automatic group, if they aren’t truly dynamic, then you’ll have to be adjusting them a lot.
If this is just to roll something out slowly, you can also divide up your computers so they aren’t relevant all at the same time window.
1 Like
We need to run software scan only in groups of 50.(This is because of the performance consideration in virtual environments) and we are looking at 2000 to 3000 new clients reporting daily and a total of 40,000+ overall clients.
Is there any way we can dynamically group clients in 50’s which are not already grouped ?
Have you thought about simply utilizing the stagger function with your bigfix action so that the scans get distributed over a period of time instead of happening all at once?
1 Like
Yes… kind of.
You can dynamically generate a certain number of “groups” but you can’t precisely determine how many machines will be in each group. If you want the number to be 50 in the groups, then you would take the total number of machines that you need to dynamically group and then divide that by 50. This would tell you how many groups you need. Then you can use relevance to distribute the machines among those “groups”, but it will not be exact.
I would recommend dynamically allocating your VMs to “groups” separately from all of your other endpoints.
There are also various methods to stagger the running of the scan throughout different times rather than just doing it by groupings.
One issue is if you have a group of 50 machines scheduled to run the scan at a particular time frame, but a few are asleep, then they may never actually run the scan. Then what?
Another concept I came up with that may work better for this is to take a month, divide it up into 2 hour chunks, and then have the computers able to run the scan in a set of different 2 hour chunks throughout the month that would be at different times of day. This would give them the opportunity to run at many different time periods randomly, but only at the interval you set, so once a week or similar.
This is also an option, but it has a maximum stagger time that is lower than I’d like, and I’m not exactly sure how well it works. I definitely need to play with it more to get a better understanding of how that could fit in.
Based on what we tested we were asked to group in 50’s because in case of virtual the VMManager(Vcenter) info is not available on Endpoint Manager and to limit the risk of all the software scan running on the same VMManger we restricted it to 50 and we can only run scans over night and not during business hours .
Can you please guide me on creating groups dynamically .
Not without more information.
How many computers do you need to do this for?
All of them are VM.
Can you please guide me on creating groups dynamically .
Hi @Manoj1
It sounds like your looking to create groups for scan purposes so your not hitting a number of VM’s that are hosted on the same blade. This is normal practice for BigFix Inventory or ILMT.
If this is your goal then the best way is to create separate scan groups for your environment and allocate VM’s to those groups using a random number.
If you have 40K endpoints, then I would suggest creating 7 groups, one for each day of the week. You can allocate a VM to one of 7 groups randomly using:
Q: (computer ID mod 7) + 1
A: 2
This gives you a random number between 1 and 7. The law of averages works very well here and you end up with a random selection of endpoints for each day of the week. Using this method you will theoretically only scan 1/7 of VM’s on any host.
Group where output = 1 would scan on Monday
Group where output = 2 would scan on Tuesday
and so forth…
3 Likes
How many endpoints/computers/VMs? You still haven’t answered this question.
I wasn’t asking how many are VMs, I was asking how many VMs are there?
If the goal is to have dynamic groups of about 50 endpoints, then I need to know how many total there are that need divided, then use a method like @GwyndafDavies mentions above to do it.
The total number of endpoints is 58k .
Then you would do something like (computer ID mod 1000) and that would give you ~58 computers per group in 1000 groups.
There are better ways to handle this dynamically while also spreading out the load.
Are all 58k endpoints VMs? It doesn’t seem like you’d have to do this for all of them.
He has indicated that they are all VM’s. It a bit further up the thread.