Title: Anthony Wiles
1TD011 STF 213 Status Report (formerly STF187)
(Jens Grabowski, Gyorgy Rethy, Ina
Schieferdecker, Anthony Wiles, Colin Willcock)
- Anthony Wiles
- MTS34
- Rome, April 2002
2TTCN-3 Edition 2
- For approval (V.2.2.0)
- DES 201 873-1 2 Core Language
- DES 201 873-2 2 Tabular Format
- DES 201 873-3 2 Graphical Format
- DES 201 873-4 1 Operational Semantics
- Issues
- Changes to Core
- Changes and Additional changes to Tabular Format
- Graphical Format (Ina)
- Part 4 - ES or TR?
- CR mechanism
- Use of TTCN-3 by TIPHON
3CRs resulting from vote
- Over 100 CRs received after extended voting
period - Mainly from Ericsson
- All resolved to agreement of STF213/Ericsson/TTCN-
3 list etc) - All changes are tracked in document CR list
- Catagories of comments
- Editorial (17)
- Test Components (2)
- String/character types/values (10)
- Type compatability etc. (12)
- Import (8)
- Regular expressions (3)
- BNF (4 TTCN-3 list)
- Test steps (1)
- ASN1 (7)
- Encoding (3)
- Miscellaneous (40)
4Changes to the Core (1) (from v.1.1.0 not just
CRs)
- Editorial/restructuring
- General editorials, corrections to examples, etc.
- Operational semantics -gt Part 4
- Predefined functions included from TS 101 875
- Clarifications
- Scope mechanism, allowed connection of PTCs
- Structural equivalence/type compatability issues
- Corrections/simplifications
- BNF
- Consitent/simpler Import mechanism
- Named alts -gt teststeps -gt altsteps (macro --gt
functional) - Timers (no negative values, zero values)
5Changes to the Core (2)
- Corrections/simplifications (cont)
- Operator precedence revised
- PTCs (e.g., scope of PTC identifiers)
- Verdicts (set.verdict -gt setverdict)
- ASN.1 issues (named bits/identifiers, mapping)
- Additions/improvements
- Regular expressions
- Power operators
- Refinement of the port model
- Return value on expressions
- Addition of superset/subset
- Any type
- Additional types (short, long, strings) by
encoding) - Ranges for float
6Changes not done
- Default encoding for TTCN-3
- Should be opened as new WI if needed
- Power operators
- Not urgent - can be added later
- IDL Interface
- Should be opened as new WI if needed
- Some work done outside ETSI
- External global timer
- Should be done in connection wit RT TTCN-3
7Early CRs on v2.2.0
- Received some CRs on v.2.2.0
- One or two BNF corrections
- Few editorials
- Keyword table param -gt parameters
- Change example for referencing nested fields
- Can approve version 2.2.0 here in Rome and fix
these points in the meantime? - Put 2.2.0 to vote meanwhile STF213 produces
2.2.1? - Produce 2.2.1 then vote? When?
8Changes to the Tabular Format (from v.1.1.0 not
just CRs)
- Some editorials
- A few corrections
- Better format/layout (not technical)
- Brought in-line with changes in Core Language
- Have received some late comments which should be
fixed - No real technical change - corrections to mapping
BNF, examples, references, couple of
clarifications - Accepted by the TFT Editor (Thomas Wernitz)
- Can approve version 2.2.0 here in Rome and fix
these points in the meantime? - Put 2.2.0 to vote meanwhile STF213 produces
2.2.1? - Produce 2.2.1 then vote? When?
9Part 3 - GFT
- Presentation by Ina
- Co-ordination with ITU-T
- OMG (Testing Profile)?
-
10Part 4 - Operational Semantics
- An ES or TR?
- Plan was to produce version 1 for approval in
April - Version 2 in October
- Go to ES twice in same year .!
- Propose to downgrade to TR for version 1 (April)
and go for ES in October - Or no approval at this meeting (go straight to
v2) - Problem - What to refer to from Part 1?
- TR (which will soon be replaced by )?
- ES (to appear)?
- Nothing?
-
11CR Mechanism
- CR Mechanism is in place
- Can be found on http//www.etsi.org/ptcc/Change_re
quest/ptccttcn3cr.htm - Direct links from MTS Portal and PTCC Website
- Defines Rules and Procedure
- Submission by CR only (downloadable template)
- Full solutions must be provided
- Mailing list TTCN3CR_at_list.etsi.fr
- Initial follow-up (record, number, status, etc) -
Gyorgy Rethy - Resolution by STF213, SG, MTS (in that order)
- Discussion (but not decision) on TTCN3 mailing
list - Final follow up
- All CRs plus Status List on web)
- Some proposal for text version of CR template
- Word is ETSI standard format
- Need Word formatted solutions, not ASCII
12Use of TTCN in EP TIPHONTM
- EP TIPHONTM
- Telecommunications and Internet protocol
Harmonization Over networks - EP TIPHON WG6 Testing and Validation
- STF196 Test specs based on TIPHONTM profiles
for - H.225 (TTCN-2)
- H.245 (TTCN-2)
- H.248 (TTCN-2)
- SIP (TTCN-3)
- OSP (TTCN-3)
- BICC (TTCN-2)
- Security - ?
- Interoperability scenarios for H.323-SIP
interworking (Text)
13What is SIP?
- Session Initiation Protocol, draft RFC2543
- IETF Internet control protocol for
- establishing, changing and tearing down voice and
multimedia calls - Like http (sipanthony.wiles_at_etsi.fr)
- Can build services like web pages
- Client-Server behaviour (simple)
- SIP Components
- Use Agents
- SIP Servers (proxy server, redirect server)
- SIP messages are text based
- Defined by a quite complex BNF
- Loose syntax
14Status of TIPHONTM SIP Conformance Test
Specifications
- IETF SIP draft RFC 2543 bis4
- 437 Test Purposes (TSSTP)
- Complete TTCN-3 ATS April 2002 (pure SIP)
- Currently validating ATS on experimental test
platform - Will be input to 10th SIPit April 2002
- Fully validated test suite May-June 2002
- Upgrade to bis9 (RFC) July-September
- Upgrade to TIPHON profile based on finalized RFC
September-December
15Issues with testing SIP (1)
- Behaviour relatively straightforward
- TTCN-3 address type and map function are useful
- SIP Messages are unstructured (ASCII text)
- Pure text gt many messages perhaps with only
small differences - Can send as pure text
- but on receive how to match?
- Control vs Readability
- Should we structure (sub-structure) each message
- Or rely on pattern-matching?
- . Or compromise?
16Issues with testing SIP (2)
- Where to place PCO
- Directly over TCP/UDP
- or allow some processing by test system
- Need some form of test adapter
- Does some form of preprocessing
- E.g., remove duplicate spaces and CRLF
- Breaks message into SIP headers (To, From, Via )
- Converts to TTCN-3 types (and vice versa)
- Mapping is defined as part of the ATS
- TTCN-3 module
- Sending of invalid or 'unusual' SIP messages done
as pure text
17Conceptual SIP Test Architecture
18A Typical SIP Message(Simplified)
SIP/2 200 OK From userBltsipxxx_at_yyy.zzzgt To
userAltsipaaa_at_bbb.cccgt CSeq 1
INVITE Content-Length 0
19SIP Message in TTCN-3(Readability vs
Controllability)
Parameterisation
template SIP_RESPONSE MyResponse (charstring
StatusCode) SIP_Version
"SIP/2.0", Status_Code StatusCode, // use
of parameter Reason_Phrase
"OK", Message_Headers ResponseHeaders, //
Substructure Message_Body omit // SDP
info. Not present template SIP_ HEADERS
ResponseHeaders // setof charstring "From
userBltsipxxx_at_yyy.zzzgt", "To
userAltsipaaa_at_bbb.cccgt", "CSeq 1
INVITE", "Content-Length "
int2char(MBLength), // All other headers are
ignored
Substructure
Setof type
Could use regexps
Function call
Wild cards
20What is OSP?
- Open Settlement Protocol, TS 101 321 V.2.1.1
- Facilitates the exchange of inter-domain
- Pricing
- Authorization
- and Settlement Information
- between Internet telephony operators.
- Works between multiple administrative domains in
a secure manner - XML- based
- ASN.1 for authorization
- Can be used with SIP or H.323
21Status of TIPHONTM OSP Conformance Test
Specifications
- TS 101 321 V.2.1.1
- 30 Test Purposes (TSSTP)
- Protocol is very open
- TIPHON profile gt more TPs
- No authorization tests yet gt more TPs
- Currently validating TTCN-3 ATS on experimental
test platform - Remote OSP server in US
- Complete June 2002
- May add new test cases in the Autumn (v2)
- E.g., authorization
- if base standard or profile develops
22Issues with testing OSP (1)
- Also simple Client - Server behaviour
- No problem for TTCN-3
- Interleaved behaviour useful (easily handles
permutations of expected behaviour) - OSP messages defined in XML
- Need to map OSP XML DTD to TTCN-3
- Not particularly difficult
- A generic mapping would be useful
- Security and Authorization (optionally) uses
ASN.1 - Not a problem for ATS
- More complex test system as it has to switch
between XML and ASN.1 encoding - Not included in this first version of ATS
23Issues with testing OSP (2)
- Where to place PCO
- Directly over TCP or UDP
- or allow some processing by test system?
- Need some form of test adapter
- Does some form of preprocessing
- Converts XML message to TTCN-3 types (and vice
versa) - Mapping is defined as part of the ATS
- Sending of invalid or 'unusual' OSP messages done
as pure XML (ASCII)
24Conceptual OSP Test Architecture
IUT SIP Implementation
TTCN-3 OSP
Adaptation Layer
HTTP
HTTP
SSL TCP
TCP
SSL TCP
TCP
IP
IP
LAN
25A Typical OSP Message in XML
lt!ELEMENT PricingIndication ( Timestamp,
SourceInfo, DestinationInfo, Currency, Amount,
Increment, Unit, Service, ValidAfter, ValidUntil
)gt lt!ATTLIST PricingIndication componentId ID
REQUIREDgt lt!ELEMENT Timestamp
(PCDATA)gt lt!ATTLIST Timestamp critical (true
false) "true"gt
26OSP Message in TTCN-3(Mapping to XML)
type record PricingIndication ID_DTD msgIdFiel
d, CDATA_DTD randField, ID_DTD compIdField,
Timestamp timestamp, SourceInfo sourceInfo, D
estinationInfo destinationInfo, Currency curren
cy, Amount amount, Increment increment, Unit
unit, Service service, ValidAfter validAf
ter, ValidUntil validUntil with encode
(msgIdField, randField, compIdField)
XML_ATTRIBUTE type record Timestamp ENUM_DT
D criticalField (true,false), PCDATA_DTD v
alueField with encode (criticalField)
XML_ATTRIBUTE
27Validation
- Have implemented experimental testing platforms
for SIP and OSP - Main task is the implementation of the adaptation
layers - Validation of SIP test cases
- Early demonstration at 7th SIPit (March 2001)
- Demonstrated at IMTC Winterop (Dec. 2001)
- Soon able to execute 400 test cases
- Will provide conformance testing 'service' at
10th SIPit (April 2002) - Validation of OSP test cases
- Achieved initial connection to remote IUT
(Transnexus OSP server) - Full validation will start in March/April
- Finish in June 2002
28Conclusions
- Specific problems when dealing with text-based
protocols - How to represent messages
- TTCN-3 has several advantages (over TTCN-2)
- Trigger
- Address type
- Map operations
- Interleaved behaviour
- Data types such as set and set of (no need to use
ASN.1) - Regular expressions
- Programming paradigm closer to datacom community
- Have implemented and executed the test suites
- So we know it works!
- Learning curve TTCN-2 -gt TTCN-3 not too high
- ETSI TBs will use TTCN-3 if there is tool (test
system) support - TIPHON, 3GPP
29Acknowledgements
- EP TIPHONTM
- Chairman Mr. Hans Van Der Veer
- WG6 Chairman Mr. Klaus Sambor
- STF 176 SIP/OSP team
- Frederique Aurouet, ACACIA (SIP)
- Csaba Koppany, Ericsson (SIP)
- Theo Vassilou-Gioles, Testing Technologies (SIP)
- Scott Moseley, FSCOM (SIP/OSP)
- Sebastian Mueller, FSCOM (OSP)
- Jean-Claude Wattelet, Wattelet Telecoms (OSP)
- Background support
- Nokia, NetTest, Alcatel, Nortel
30