2 of 2 people found this helpful
You can use the VA data in a correlation rule in the following way - rules 146-10 and 306-10 in the Policy Manager > ESM Rules will trigger if the asset identified by either the source or destination of the event may be susceptible to a known vulnerability. You can use those signature ID’s as part of a correlation rule or as part of an ACE Risk Correlation Rule by filtering on those specific event id's.
The other main use case, other than an alert, for VA information is using the ESM Views to report and organise the VA information. It would be interesting to hear if there are some users out there in the community who might have some other ideas or uses for that information.
To me, the best thing for VA is automate asset creation with their information.
Would you please elaborate a little bit more. What do you mean by automating asset creation using VA data.
Actually, it's nothing much. What I try to say is you don't have to create assets on SIEM manually. If you can connect SIEM with VA source then it allowed SIEM to retrive information from VA source. As a result, assets on SIEM will be automatically created for you with basic information such as their OS version and port opened for example. Then you can group those assets into lists and later use lists in correlation rule.
I just look at your question again and notice that you refer to vulnerability assessment data. So this may be not related. You should know about VA source already, I guess. Please ignore my messages.
Using the information provided by you, can we possibly setup a correlation rule that checks for
1-> Any attack comming on an asset XYZ detected by the IDS/IPS
2-> The asset XYZ has a known vulnerability as reported by Vulneraility assessment tool
If both conditions are met then generate an alert. Both IDS/IPS and VA tool data are in the SIEM. However we are not using Nitro IPS but instead are using IBM IPS.
I think that the second condition in above rule is met by ESM rules 306-10 and 146-10 if I understood you correctly. However, I have no idea how to setup the first condition in a correlation rule. Please shed some light on this.
Thanks and Regards
You are correct that the second condition will be met by those ESM rules and if you are using an IBM IPS then that alert will not trigger. However, there is a way of doing this but its not very ganular so you would get some false positives. Also we have been talking internally about this and there are enhancements for this type of scenario so stay tuned for further releases of the product.
What you can do currently is select your IPS in the System Tree and then Event Views > Normalized Event Summary. See how your IPS events are normalized - you could try adding this as a filter for Normalized ID 536870912/5 which is for Exploit. I dont have access to another vendor IPS so I cant give you any more specific information. What we are trying to do here is find how the IPS events, which are vulnerability exploits, are categorized.
Once you have that information then you can create a correlation rule triggering on the Normalized Rule(s) for those vulnerability exploits and Asset or Asset Group. In addition you can add the Device ID of your IPS as shown below.
I hope this helps.
Thank you for a detailed reply. I talked to our IPS guy and the reponse I received was that "All events logged in the IPS are exploit events" as all signatures in the IPS are basically for a specific attack type or vulnerability. Please correct me if I am wrong.
So as you said that "What we are trying to do here is find how the IPS events, which are vulnerability exploits, are categorized" seemed not possible as all events in IPS are possible vulnerability exploits.
Moreover, could you please tell what is the result of the condition "Device ID (In) [A_IPS]" ? Does it return all events that are logged in the IPS? I created a rule having this one condition only and the resultant events were some random IPS events not all of them. Kindly explain what it is actually.
What's your opinion on the following correlation rule?
1) Signature ID in 306-10, 146-10
2) Device ID in IPS (further fine tune this by e.g neglecting event sub types that are blocked/droped etc)
I only have access to the McAfee IPS and the way we normalize those events in quite granular. So for example, if the IPS detects limewire that is normalized under Policy > P2P Policy > Limewire. As another example if there is a Shellcode detection event seen by the IPS then we normalize that in Exploit > Shell Script Execution > Shellcode detection. As you can see, for our IPS not all events are normalized as Expolits. My suggestion for "filtering" on that in a correlation rule was to try and minimize the number of alarms to IPS events which are trying to exploit a specific vulnerability. However, this may not be possible on your IPS and would depend on how the events are normalized.
On my test system I setup a DataSource for a thirdparty IPS and named it A_IPS. So the idea of adding that condition is to limit the exploit events to events from that specific DataSource.
The correlation rule signatures 306-10 and 146-10 are triggered on the ESM when an Asset is vulnerable to an event but those will not trigger on events from an IBM IPS. So I still think that the previously suggested correlation rule filter, or variation on that, would be the best option.
Thanks for the clarification.