260 lines
No EOL
6 KiB
Markdown
260 lines
No EOL
6 KiB
Markdown
TOP SECRET STRAP 1
|
||
|
||
SMO
|
||
Better Analysis
|
||
MTI Mastering the Internet
|
||
Business Change
|
||
|
||
# Next Generation Events
|
||
|
||
23 March 2009
|
||
|
||
TOP SECRET STRAP 1
|
||
|
||
---
|
||
|
||
TOP SECRET STRAP 1
|
||
|
||
## What is NGE?
|
||
|
||
* Systems like HAUSTORIUM reaching ingest capacity
|
||
* But scale and variety both increasing
|
||
|
||
* 5-Eyes also far-apart on “metadata” requirements, need to get closer together
|
||
|
||
The Answer?
|
||
|
||
* NGE: A multi-stage project that tackles a series of the problems, at increasing scale, and with increasing collaboration
|
||
|
||
29 September 2009
|
||
|
||
---
|
||
|
||
TOP SECRET STRAP 1
|
||
|
||
## Next Gen Events: High-level Plan
|
||
|
||
53 x 10G
|
||
100 x 10G
|
||
200 x 10G
|
||
|
||
Initial Cloud installed
|
||
|
||
Initial Cloud Ops Capabilities Delivered in CDW
|
||
|
||
Large Scale Events Capability Available
|
||
|
||
Jan
|
||
Jan 11
|
||
Jan 12
|
||
|
||
BLAZING SADDLES Complete
|
||
|
||
Convergence Capabilities Available in CDW
|
||
|
||
Enhanced Internet Profiling Capabilities Available
|
||
|
||
Data Mining and Large Scale Contact Chaining capability available
|
||
|
||
Experimental Capability
|
||
|
||
Scaled Capabilty
|
||
Full Operational Useage
|
||
|
||
Q4 2009
|
||
- NSA QFD Sharing
|
||
- Initial Viz
|
||
- Converged QFDs
|
||
- Dev Cloud
|
||
- ?ECB
|
||
|
||
Q1 2010
|
||
- TEAMSOC1-4 (JCE)
|
||
- More QFDs Available
|
||
- CDW Con-op
|
||
- Viz and services
|
||
|
||
Q2 2010
|
||
- Initial CDW Cloud
|
||
- More QFDs available
|
||
- Cloud Analytics
|
||
- Viz and services
|
||
|
||
Q3 2010
|
||
- Viz and Services complete
|
||
- Strategic Sharing
|
||
- Bude refurbishment
|
||
- ? BSS Collaboration
|
||
|
||
Q1 2011
|
||
- Large Scale CCT
|
||
- Data Mining
|
||
- Large Scale Cloud
|
||
- BAU Transition
|
||
- ? Decommissioning
|
||
|
||
New technologies (particularly from JCE) incorporated into solution as they are de-risked / proven
|
||
We Are Here
|
||
|
||
29 September 2009
|
||
|
||
---
|
||
|
||
TOP SECRET STRAP 1
|
||
|
||
## NGE: The Last Three Months
|
||
|
||
* Sharing Enriched Metadata (HARBOUR PILOT)
|
||
* Moving towards metadata standards across 5-Eyes
|
||
* Invisible to GCHQ analysts
|
||
|
||
* Internet Profiling (BLAZING SADDLES)
|
||
* Taking ICTR ideas on how to process Events at scale, and scale even more
|
||
* Required significant effort on End-to-End Sigint process
|
||
|
||
29 September 2009
|
||
|
||
---
|
||
|
||
TOP SECRET STRAP 1
|
||
|
||
## Plug 1 - Internet Profiling: The BLAZING SADDLES Delivery
|
||
|
||
* What It Does:
|
||
* Takes 8 ICTR QFD's and scales them for up to 100 x 10G bearers
|
||
* Allows the analyst to see large amounts of a targets online activity
|
||
* Metadata – MUTANT BROTH, AUTO ASSOC, KARMA POLICE, SOCIAL ANIMAL, INFINITE MONKEYS, HRMAP
|
||
* Content - MEMORY HOLE, MARBLED GECKO
|
||
|
||
* Why You Care:
|
||
* Want to know alternate online accounts?
|
||
* Quickly build up a picture of someone's online MO and interests?
|
||
* Identify for further exploitation (with other techniques) a targets network/machines?
|
||
* Success across IP/X – CP, SIMMER, Mumbai, G20 – and ask around in your IPT!
|
||
|
||
* How You Get Access:
|
||
* Currently instigating corporate process (based on C2C skill level)
|
||
* Interim – see your Tech Director or Tech Ex
|
||
|
||
29 September 2009
|
||
|
||
---
|
||
|
||
TOP SECRET STRAP 1
|
||
|
||
## NGE: The Next Three Months
|
||
|
||
* ROCK RIDGE
|
||
* Continuing QFD roll-out
|
||
* SAMUEL PEPYS
|
||
* CAFFEINE HIT
|
||
* Sharing some QFD's with (initially) NSA
|
||
|
||
* Converged Events
|
||
* Ensuring we don't perpetuate the C2C/Telephony divide
|
||
* Specific QFD's that enhance our ability to exploit converged
|
||
* Evolved MUTANT BROTH
|
||
* LAUGHING HYENA
|
||
* Exit strategy for SALAMANCA/HAUSTORIUM
|
||
|
||
* CLOUD Experiments at Bude
|
||
* JCE and TINT
|
||
* Developing/testing technologies for later in the roadmap
|
||
|
||
* ICTR (and others!) continue to develop new ideas
|
||
|
||
29 September 2009
|
||
|
||
---
|
||
|
||
TOP SECRET STRAP 1
|
||
|
||
## NGE: And After That?
|
||
|
||
* Capability Development Workspace
|
||
* Bulk datamining capability
|
||
* Use existing sources, and new cloud capabilities
|
||
|
||
* Large-scale contact chaining
|
||
* MOAG – but anyone can create
|
||
* Using both GCHQ and NSA datastores
|
||
|
||
* MO/Profiling based discovery
|
||
* Always been the goal for events-led analysis
|
||
* Dependent on technological advancements, but looking good
|
||
|
||
* Events/Content Fusion & Visualisation
|
||
* Seamless navigation between Events and Content
|
||
* Making sure we continue the MONTE VISTA/LOOKING GLASS ideas
|
||
|
||
29 September 2009
|
||
|
||
---
|
||
|
||
TOP SECRET STRAP 1
|
||
|
||
## Next Generation Content..?
|
||
|
||
* Not yet...but thinking and delivery is happening
|
||
* TIPC using TDI's
|
||
* Expand XKS use
|
||
* Trial new ways of collecting/processing content (TINT)
|
||
|
||
29 September 2009
|
||
|
||
---
|
||
|
||
TOP SECRET STRAP 1
|
||
|
||
## Plug 2: TIPC Expansion
|
||
|
||
* What It Does
|
||
* Full client IP stream collection triggered by known selector
|
||
* Expanded to STM-64 environment as well as STM-1/4
|
||
* Now triggered by TDI's, not just gmail, yahoo and maktoob
|
||
|
||
* Why You Care
|
||
* Unique Intelligence material that can't be strong selected – web visits/searches etc
|
||
* Find new protocols used by targets – you, tech trends, T development
|
||
* Contextless
|
||
* New dictionary – old one completely erased
|
||
|
||
* How Do You Get Access?
|
||
* Talk to your C2C Tech Ex – they are running pre-requisite briefings as there are some dangers...(full IIB!)
|
||
|
||
29 September 2009
|
||
|
||
---
|
||
|
||
TOP SECRET STRAP 1
|
||
|
||
## Plug 3: XKS & TINT @Bude Experiments
|
||
|
||
* What It Will Do:
|
||
* Promotion from XKS to IIB
|
||
* Integration into LOOKING GLASS
|
||
* Connection to Native File Viewer (FUME CUPBOARD)
|
||
* Continuing to work on the NSA data access issue
|
||
* The TINT@Bude Experiments Attempt To:
|
||
* Re-sessionise everything
|
||
* Tag traffic, based on
|
||
* strong selector/geography/ application
|
||
* contextual fingerprints:
|
||
* Extract metadata in bulk
|
||
* Retain a 3-day rolling buffer of ‘interesting' content
|
||
* for retrospective/protocol/network/analysis
|
||
* for refining fingerprints/selectors
|
||
* Do this on 20 x 10G's!
|
||
|
||
* Why You Care:
|
||
* Packet processing approach misses stuff
|
||
* Strong selection only
|
||
* Too much data retained is unused (97% unviewed)
|
||
* Promote only the good stuff to long-term storage
|
||
* Aim: to automatically promote to long term storage
|
||
|
||
* When Do You Get Access?
|
||
* New XKS capabilities will be rolled out to GCHQ KS's when available
|
||
* TINT PUT in place, but experimental, not operational use only
|
||
|
||
29 September 2009 |