[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [cti-cybox] Re: [Non-DoD Source] RE: [cti-cybox] Network Connection Object Refactoring
Why not just copy ipfix (net flow) structure. Its already a standard way of capture flow records including ability for extension fields necessary to capture all relevant information for a flow.
I don’t see why this needs to be re-invented.
allan Thomson
LookingGlass CTO
> On Feb 16, 2016, at 6:58 AM, Kirillov, Ivan A. <ikirillov@mitre.org> wrote:
>
> So if we’re going towards support of all OSI layers, what would be the set of attributes common to all, that would be part of the “base” Network Connection Object? I’m thinking it’s probably just Start_Time and End_Time.
>
>
> Anyhow, this would give us something like:
>
> Network Connection Object
> * Start_Time
> * End_Time
> * Layer 1 extensions
> * Ethernet physical
> * Bluetooth
> * Layer 2 extensions
> * Ethernet
> * Layer 3 extensions
> * IP
>
> * Layer 4 extensions
> * TCP
> * UDP
> * Layer 7 extensions
> * HTTP
> * DNS
>
> Regards,
> Ivan
>
> On 2/16/16, 7:39 AM, "Casey, Eoghan CIV DC3/DCCI" <Eoghan.Casey@dc3.mil> wrote:
>
>> Jason,
>>
>> We are proposing a refactoring of the Network_Connection object for exactly the reasons you state. The proposed refactoring does not assume IP-based networking, but must fully support this common protocol. You will observe that the proposed refactoring is designed with encapsulation in mind, and can represent Layer 1.
>>
>> I encourage you to contribute an example along the lines you describe to see if you find any gaps/weaknesses in the proposed refactoring that require filling/strengthening.
>>
>> Eoghan Casey
>> Chief Scientist
>> Defense Cyber Crime Center (DC3)
>> 410-694-4329
>> Eoghan.Casey@dc3.mil
>>
>> -----Original Message-----
>> From: cti-cybox@lists.oasis-open.org [mailto:cti-cybox@lists.oasis-open.org] On Behalf Of Jason Keirstead
>> Sent: Tuesday, February 16, 2016 7:42 AM
>> To: Terry MacDonald
>> Cc: Kirillov, Ivan A.; cti-cybox@lists.oasis-open.org
>> Subject: [Non-DoD Source] RE: [cti-cybox] Network Connection Object Refactoring
>>
>> You are making a big assumption there that the network connection is point-to-point and is using the IP protocol. Not all network protocols are point-to-point, many are broadcast or mesh.
>>
>> I made some related comments on the doc - but my exec summary of this is, we need to stop thinking only about IP based networking. We have to be able to model CTI that happens over any protocol, For example what about an emerging threat that uses Zigbee, how would I model that using this proposal? It would not align at all.
>>
>> I believe the network object should start at Layer 1 and each protocol layer be extensions on top of that. Having assumptions that you are starting at Layer 3 (IP) is not a good starting place, it is way too limiting, and will result in us later on having to craft things like "ZigbeeConnection" objects because we can't use the existing network connection.
>>
>> -
>> Jason Keirstead
>> STSM, Product Architect, Security Intelligence, IBM Security Systems www.ibm.com/security | www.securityintelligence.com
>>
>> Without data, all you are is just another person with an opinion - Unknown
>>
>>
>> Inactive hide details for Terry MacDonald ---02/16/2016 05:59:35 AM---A question.... a network connection is a communication reTerry MacDonald ---02/16/2016 05:59:35 AM---A question.... a network connection is a communication relationship between an initiator and a recip
>>
>> From: Terry MacDonald <terry@soltra.com>
>> To: "Kirillov, Ivan A." <ikirillov@mitre.org>, Jason Keirstead/CanEast/IBM@IBMCA
>> Cc: "cti-cybox@lists.oasis-open.org" <cti-cybox@lists.oasis-open.org>
>> Date: 02/16/2016 05:59 AM
>> Subject: RE: [cti-cybox] Network Connection Object Refactoring Sent by: <cti-cybox@lists.oasis-open.org>
>>
>> ________________________________
>>
>>
>>
>>
>> A question.... a network connection is a communication relationship between an initiator and a recipient. The initiator service opens a network connection and is given a source port sends traffic from an IP address to the recipient. The recipient service has separately created a listening port bound to one or more IP addresses on the listening device. The network connection is a connection between those two items at a particular time.
>>
>> My question is… shouldn’t we represent a network connection as an initiator endpoint object and a recipient endpoint object and a connection object showing the communication between them as a connection between the IP/port combinations at either end?
>>
>> In other words:
>>
>> Initiating Service resides at 172.16.12.34 on TCP Port 24356, Recipient Service resides at 10.16.1.44 on TCP Port 80, Network connection is communication at a particular time between these two services.
>>
>> This is slightly different to the proposal https://docs.google.com/document/d/1k5cZiiOgo4WjXN2GLRZ0-1vO8wstKnPZLbjwlb23Hkg/edit# <https://docs.google.com/document/d/1k5cZiiOgo4WjXN2GLRZ0-1vO8wstKnPZLbjwlb23Hkg/edit#> , as it would treat the endpoints as separate things, and would relate the two entities together independently
>>
>> "objects": [
>> {
>> "id": "object--1",
>> "type": "NetworkConnection",
>> "service": "tcp",
>> "start_time": "Jan 4, 2015 10:00",
>> "end_time": "Jan 4, 2015 10:15",
>> "extended-properties": {
>> "state": {
>> "connection_state": "sf",
>> "overall_state": "established"
>> },
>> "packet": {
>> "initiator_packets": 14356,
>> "recipient_packets": 14326
>> "initiator_bytes": 35779,
>> "recipient_bytes": 935750,
>> }
>> }
>> },
>> {
>> "id": "object--2",
>> "type": "NetworkedService",
>> "extended-properties": {
>> "ipv4": {
>> "address": "172.16.12.34"
>> }
>> "tcp": {
>> "port": "24356"
>> }
>> },
>> {
>> "id": "object--3",
>> "type": " NetworkedService",
>> "extended-properties": {
>> "ipv4": {
>> "address": "10.16.1.44"
>> },
>> "tcp": {
>> "port": "80"
>> }
>> }
>> }
>>
>> ],
>> "relationships": [
>> {
>> "id": "relationship--1",
>> "type": "relationship",
>> "from": "object--1",
>> "to": "object--2",
>> "relationship_value": "initiated_connection_from"
>> },
>> {
>> "id": "relationship--2",
>> "type": "relationship",
>> "from": "object--1",
>> "to": "object--3",
>> "relationship_value": "received_connection_on"
>> }
>> ]
>> }
>>
>> Comments?
>>
>> Terry MacDonald
>> Senior STIX Subject Matter Expert
>> SOLTRA | An FS-ISAC and DTCC Company
>> +61 (407) 203 206 | terry@soltra.com <mailto:terry@soltra.com>
>>
>>
>> From: cti-cybox@lists.oasis-open.org [mailto:cti-cybox@lists.oasis-open.org] On Behalf Of Kirillov, Ivan A.
>> Sent: Tuesday, 16 February 2016 11:09 AM
>> To: Jason Keirstead <Jason.Keirstead@ca.ibm.com>
>> Cc: cti-cybox@lists.oasis-open.org
>> Subject: Re: [cti-cybox] Network Connection Object Refactoring
>>
>> Sure, here it is: https://docs.google.com/document/d/1k5cZiiOgo4WjXN2GLRZ0-1vO8wstKnPZLbjwlb23Hkg/edit?usp=sharing <https://docs.google.com/document/d/1k5cZiiOgo4WjXN2GLRZ0-1vO8wstKnPZLbjwlb23Hkg/edit?usp=sharing>
>>
>> Please add any comments there; I know I definitely missed some as you had mentioned. Oh, and pardon any formatting issues - this was a direct copy of the rendered markdown.
>>
>> Regards,
>> Ivan
>>
>> From: Jason Keirstead <Jason.Keirstead@ca.ibm.com <mailto:Jason.Keirstead@ca.ibm.com> >
>> Date: Monday, February 15, 2016 at 4:55 PM
>> To: Ivan Kirillov <ikirillov@mitre.org <mailto:ikirillov@mitre.org> >
>> Cc: "cti-cybox@lists.oasis-open.org <mailto:cti-cybox@lists.oasis-open.org> " <cti-cybox@lists.oasis-open.org <mailto:cti-cybox@lists.oasis-open.org> >
>> Subject: Re: [cti-cybox] Network Connection Object Refactoring
>>
>>
>> There was a substantial amount of feedback on this proposal on Slack a few weeks ago... many of which aren't captured.
>>
>> Can this be re-posted somewhere we can comment directly on it, such as Google Docs? It would greatly help the discussion.
>>
>> I don't really want to re-hash all of my slack chat comments over email (or on the phone)... but I can do that if needed...
>>
>> -
>> Jason Keirstead
>> STSM, Product Architect, Security Intelligence, IBM Security Systems www.ibm.com/security <http://www.ibm.com/security> | www.securityintelligence.com <http://www.securityintelligence.com/>
>>
>> Without data, all you are is just another person with an opinion - Unknown
>>
>>
>> Inactive hide details for "Kirillov, Ivan A." ---02/15/2016 05:35:27 PM---All, Here is a community contributed proposal around "Kirillov, Ivan A." ---02/15/2016 05:35:27 PM---All, Here is a community contributed proposal around refactoring the Network Connection Object and r
>>
>> From: "Kirillov, Ivan A." <ikirillov@mitre.org <mailto:ikirillov@mitre.org> >
>> To: "cti-cybox@lists.oasis-open.org <mailto:cti-cybox@lists.oasis-open.org> " <cti-cybox@lists.oasis-open.org <mailto:cti-cybox@lists.oasis-open.org> >
>> Date: 02/15/2016 05:35 PM
>> Subject: [cti-cybox] Network Connection Object Refactoring Sent by: <cti-cybox@lists.oasis-open.org <mailto:cti-cybox@lists.oasis-open.org> >
>>
>> ________________________________
>>
>>
>>
>>
>>
>> All,
>>
>> Here is a community contributed proposal around refactoring the Network Connection Object and related Objects: https://github.com/CybOXProject/schemas/wiki/CybOX-3.0:-Network-Connection-Object-Refactoring <https://github.com/CybOXProject/schemas/wiki/CybOX-3.0:-Network-Connection-Object-Refactoring>
>>
>> The main points around this refactoring are:
>>
>> 1. The hierarchy around Network Objects would be replaced with an extension-based approach (as with the File Object) that revolves around the “base” Network Connection Object
>>
>> 1. A base set of properties common to all network connections would be defined
>> 2. Existing layer 7 Objects (HTTP Session and DNS Query) would be become extensions
>> 3. The Network Flow Object would likely be split up into its components (e.g, YAF log, Netflow log, etc.), each of which would be an extension
>> 4. New extensions for common Network Connection properties, e.g., port, state and packet statistics, would be added
>>
>> 2. Connections to destination/source IP addresses would be defined via relationships
>> 3. The existing Socket Address Object would be deprecated, as given the specification of IP addresses via relationships and the new port Network Connection Object extension it will no longer be necessary
>>
>> We plan on discussing this (at least these high-level points) during tomorrow’s working session.
>>
>> Regards,
>> Ivan
>>
>>
>>
>>
Attachment:
signature.asc
Description: Message signed with OpenPGP using GPGMail
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]