Forum Discussion
Ryan Heffernan
Microsoft
Feb 22, 2019Join Our Azure Sentinel Community
Visit Our Blog
Now that we have announced Azure Sentinel, we'd like to invite you to speak directly to our engineering team. We believe that the best way to improve our products is by having ...
- Mar 28, 2019
Please continue providing feedback here on the Azure Sentinel Communities, if you're specifically asking for a feature request on a product go here.
https://feedback.azure.com/forums/920458-azure-sentinel
referencing Ryan's Community post : https://techcommunity.microsoft.com/t5/Azure-Advanced-Threat-Protection/Join-Our-Security-Community/m-p/311170
"
We want you to speak directly to our engineering teams. We believe that the best way to improve our security products is by having no barriers between you and the people that create them. That's why we need your participation in our security community.
As part of our community you can influence our products and get early access to changes by participating in private previews, giving feedback, requesting features, reviewing product roadmaps, joining webinars and calls, or attending in-person events.
Join Us
To join our community, click here, and then click the join button and the heart icons of the groups your are interested in, as pictured below.
"
AndrewX
Jun 20, 2019Iron Contributor
Thanks for these excellent tooling. We enable and integrated with AAD, O365, Security Center day we heard about it and very cool so far.. I am still a bit muddy though on how my existing non-azure oms agent hosts that currently send data to an existing log analytics workspace, and how the agents gets data to ATP, Security Centre send data to Sentinel. What's the best practice for agent install on hosts to get data to all the security portals?
- Chris BoehmJun 20, 2019
Microsoft
I believe this comes down to where you're needing the data, the OMS agent can be multihomed
https://blogs.technet.microsoft.com/msoms/2016/05/26/oms-log-analytics-agent-multi-homing-support/
This allows you to send data to multiple different workspaces. Be aware you'll be charged twice for the data.
If you're wanting to take advantage of the services you're already paying for you should have something like this, I'm going to be using Azure Security Center as an example.
Server -> MMA/OMS Agent--> Azure Security Center --> Azure Sentinel
This way you'll still have all the data within Azure Security Center's Workspace, you'll get security related alerts ingested into Azure Sentinel.
You can take another approach as to having Azure Sentinel and Azure Security Center together by using the same workspace.
Server -> MMA/OMS Agent -> Workspace(Azure Security Center/Azure Sentinel)
You'll see a lot more raw events this way, get Azure Security Center benefits within the same workspace, but still able to use the investigation/alerts/automation with Azure Sentinel with the additional information.
Hope this helped answer your question
- David CaddickJun 20, 2019Iron Contributor
Chris Boehm @Andrew Huddleston
From what I can make out:
Server -> MMA/OMS Agent--> Azure Security Center --> Azure Sentinel
This would mean the data/alerts/logs would end up in *both* the ASC & the Sentinel instance?
Doubling the data costings? this may/may not be desirable?
Server -> MMA/OMS Agent -> Workspace(Azure Security Center/Azure Sentinel)
This keeps it more simple technically + single cost of data/alerts/logs storage based on retention settings.
Is this essentially correct Chris?
This seems to get to the heart of the matter: although it seems to be from a while back?
Determine the number of workspaces you need
- David CaddickJun 20, 2019Iron Contributor
Hi Chris,
Chris Boehm @Ofer Shezaf
Can you confirm the details here still hold true?
https://github.com/uglide/azure-content/blob/master/articles/log-analytics/log-analytics-manage-access.md This seems to be a very good "simple" primer to start with?
It also seems to be mirrored here:
https://docs.microsoft.com/en-us/azure/azure-monitor/platform/manage-access
I guess our concerns are - we need to max the retention to 730 days (2 Years, although 3Y would be nice) - and with that we trying to ensure that we don't accidently "double-down" on the same Logs in more than one place?
"Best Practices" for Log Analytics - is it a case for just using a single Workspace for everything?
What about different data being retained for different lengths of time for instance?
Question - Can Sentinel just make a "call" to reference other Workspaces?
Obviously this would not be ideal from a performance/latency point of view...
*OR* is it a case that the only way Sentinel can run effectively is to ingest that Log data *again*?
Looking at the first screen for Sentinel - you can potentially connect a number of Workspaces?
But once you "step in" to the Console as such the Schema only lists the current Workspace.
So this appears to rule out the idea of having multiple Workspaces?
Hoping this all makes sense? ;-)
- Sankarasubramanian ParameswaranJun 20, 2019Iron Contributor
how we cam enable azure senitel
- Chris BoehmJun 20, 2019
Microsoft
Sankarasubramanian Parameswaran
Please view this link to getting started with Azure Sentinel: https://docs.microsoft.com/en-us/azure/sentinel/quickstart-onboard
It's already in open preview.