BigFix design questions

Usage and Config Platform
1

(imported topic written by SystemAdmin)

Hello,

We are in the planning phase of migrating to BigFix(most probably Systems LifeCycle Manager) from Tivoli Provisioning Manager for Software(IBM’s software dist/inv/patch management).

We have about 100 branches with 2 to 3 pc’s per a branch and no servers in the branch(the network between the head office and the branches is an MPLS cloud). Currently with the Tivoli solution we use a feature called peering that allows the pc’s in the branch to pull the package from their peers that have already received the package. Is there similar functionality in BigFix so that we do not consume all the bandwidth trying to send the package to all 3 machines in a branch together?

Regards

Pritesh

2

(imported comment written by BenKus)

Hi Pritesh,

Welcome to BigFix and we hope you have an excellent migration process.

In BigFix, we use a concept called a “Relay” that is similar to the “peering” model. Basically it will work like this:

  1. Install the BigFix Agent on your computers in your branches.

  2. Designate one computer in each branch a “Relay”.

  3. Set all the computers to automatically find their closest Relay.

Once you do that, each branch office computer will automatically report to the relay in its office (so it will use the relay for control traffic and for downloading packages). Here is some information on relays:

http://support.bigfix.com/bes/misc/besrelays.html

Additionally, you might also consider setting up bandwidth throttling for the relays themselves (or for agents in the branches if you don’t want to install relays). More information on bandwidth throttling here:

http://support.bigfix.com/bes/misc/besthrottling.html (note that there are Tasks in the BES Support Fixlet site that help you set up bandwidth throttling).

Ben

3

(imported comment written by SystemAdmin)

Hi Ben,

I spent some time looking at the links you provided. It looks like the relay performs a similar function to the depot in the TPMfSW product. The problem I will face with making one of the pc’s a relay is that if the pc is switched off by the user then the other pc’s in the branch will either not be able to receive the software or I could configure them to connect to a different relay or even directly to the BES server but this could start getting out of control too.

Where can I find more detailed info on configuring the relays and bandwidth throttling.

Regards

Pritesh

4

(imported comment written by NoahSalzman)

Pritesh,

By default a client will attempt to reach the server’s relay if the local relay is down. However, you can create your own

affiliation rules

to manage the hierarchy of relays.

Start here

to learn about bandwidth throttling.

5

(imported comment written by BenKus)

Hi Pritesh,

Typically with small branch environments like yours, we would recommend the following:

  • Create a relay in the central datacenter to service all the branches.
  • Create a relay in the branch environments that have multiple computers… If there are only 2-3 computers, you might consider skipping this step…
  • Enable static bandwidth throttling for the agents in the branches to ensure that even if the branch relay goes down, you won’t overwhelm the bandwidth.
  • Turn on relay autoselection.
  • (You might also consider setting the central relay as the “Failover relay” for the agents")

With this method, the agents will find and report to the local relay if it is available. If the relay is not there or if it is powered-off, then no problem because the agent will just download the files from the central relay at a throttled rate.

For any branches that have many computers (over 5), then you almost certainly will want a relay.

Ben