Teast team minutes 12/11/96

ICC Test Committee Meeting - 12/11/96

Agenda:

I. Introductions

II. Test Plan review

A. Overview of configurations

B. Review of table of contents of test plan

C. new tests to add

D. Timing testing (pot dial delay)

III. SCCP Looping Discussion

IV. Operator Services

V. NPAC Input for test plan/support

VI. Other Issues

I. Introductions:

Attendees:

Bill Belshaw, MCI

Ralph J. Brown, AT&T

Rich Carter, Lockheed

Andy Chalk, Sprint

Joe Clark, Bellcore

Carmen Colella, Ameritech

John Good, Ameritech

Ed Hendrix, BellSouth

OC Jackson, AT&T

Gene Johnston, GTE

Robert Jungemann, GTE

Larry Lovett, Sprint

Steve Markowski, Lockheed

John F. Shea, JFS Telecom Consulting (Lockheed)

Bettie Shelby, MFS

Walt Subora, Ameritech

Allena Wheatley, BellSouth

II. Test Plan review

Object of today is to come consensus on draft test plan provided by Bill

Belsahw. Walt handed out disks to those who had not been able to upload from

http://www.ported.com/.

A. Overview of configurations.

1. Need at least one P->P call that is intra-switch. AT&T

will add to the local section.

2. Discussion on Trunk Sub Group/Signal Ported Number

option.

Walt drew a diagram of what the Signal Ported Number

option is. It is basically the ability of a tandem to

route calls to a Ported Number on a Non-LRN capable

switch. The Tandem, when it sees this option,

treats the call as if it has encountered an MF route,

pulls the DN out of the GAP and puts it in the Called

Number parameter, sets the FCI bit to Not queried (if

SS7), and terminates the call to the switch.

The calls involving signal ported number can be covered

by the "N2" configurations. The "N3" configurations can

be deleted from the configuration matrix.

The question of Non-conforming end office test cases to

be added was discussed and it was decided that the SPN

tests took care of the termination cases and the

origination s were already in the test plan.

B. Review of Test Plan index

1. Question: any AIN tests included? So far, AIN has not

been considered in the inter-company tests. Routing

numbers can be ported DNs.

Discussion centered around interactions with triggers.

Ameritech will submit test cases for this section (4.5.8)

to Bill for inclusion in the test plan.

2. For Calling Name delivery, MWI, LIDB, and CLASS

subsytems, we need test cases to cover the situation

where the 10-digit LNP GTT result is "null". Need to

characterize what happens when the query fails

and a time-out occurs.

Bettie (MFS) said that she could probably put an office

into the testing mix that is in that condition.

3. Message waiting indicator discussion

New LEC provides MWI from old VMS. That's covered.

What about the case when the old VMS and the old location

of the client were in different LEC networks to start

with? With 30+ voice mail service providers in Chicago,

not all on Ameritech switches.

Ameritech promised to deliver one or two new test plans

to cover this kind of case.

4. Bill (MCI) added cases for operations. They are generic.

Operations flows were decided by the operations group,

but have been re-opened. Until the issues are resolved,

issue 1 will only reflect the generic cases. Will be re-

written when the flows are again agreed to and stable in

issue 2.

D. Delay Tests

Important to do something for post-dial delay, but this

is a very complex subject. It may be cause unnecessary

confusion if the job isn't done right. Will be discussed

for issue 2.

E. As of today the LSMS test section is empty. Needs to be

covered in that area, too. FCC requires overall

characterization of all tests that have been done.

III. SCCP looping problem.

The SCP Subcommittee submitted a request for this group

to cover SCCP looping as part of the test plan.

SCCP loping can occur in these ways:

(1) for a short time during porting while there is a

mismatch. Test would verify that trouble goes away

by itself when the data catches up.

(2) data out of sync long-term. Test would verify that

the error is detected and corrective action takes

place.

(3) possibly other reasons, not well-defined.

No consensus on when and how the GTT data will be checked

for mismatches among different companies. The need for

such checks is clear. Process flows need to include

whatever audits are to be done. These would then be

included in the repair service test cases.

Walt will send copies of proposed schemes to detect and

block SCCP message looping. Translation type mapping and

SCCP hop-counter.

We can discuss further at January meeting.

IV. Operator Services

Handout from Carmen Colella, AIT. Asking for tests from every network that will

use the Ameritech LIDB. Since a collect or third-party-billed call can come

from anywhere, the LIDB queries will go to and from all possible networks. This

leads to national compatibility issues.

Test plan layout shown is exhaustive; Lot's of combinations.

Carmen suggested a separate committee for testing of operator services. There

was a requirements OS subgroup. They would be a logical place to start to

constitute such a committee. Barry will check with operations team to get some

names. Could include SMEs from other than the direct participants on this

committee.

The Statement "LIDB validation should work" is not sufficiently specific as a

test result.

Asking the operator services team to say what the operator services test plan

ought to look like.

For now, freeze what we have for Issue 1. Address updates from the new team in

Issue 2 and later.

V. NPAC Input for test plan/support

1. Discussion of how NPAC would be incorporated.

Most NPAC test plans are currently proprietary. Since

they can't be incorporated in the public test plan, a

characterization of the coverage is necessary in the

master test plan, to show that it has been covered.

Sect 4.2.4 of test plan will have a matrix or summary

which characterizes the coverage. Then, at the

conclusion of testing, Lockheed would have to provide

documentation of the test results.

For Issue 1, sections on NPAC, LSMS, SCP could include an

assertion that these areas will be provided in Issue 2.

Note that provisioning tests will exercise these elements

and therefore the provisioning test results should be

provided as results for these specific sections.

2. Lockheed gave overview of test plan/schedule.

Lockheed test lab in Bridgewater, NJ. Capable of

stack-to-stack tests, security tests, managed-object

conformance tests, simulated application-to-application

tests.

Turn-up Qualification of SMSs against NPAC is required.

Expecting to get much of that from the live testing next

spring. Turn-up test plan still in preparation. Mostly

a subset of internal development testing.

Once in live service, there will be a live test bed, like

the Chicago test bed. Available to newcomers and those

making changes.

As a contingency, Lockheed promised to arrange for

simulators if the are insurmountable schedule problems

at testing time.

Lockheed handed out a proposed schedule using 3 locations

to support turn up testing.

___________________________________________________________

|Chicago |ATT |ATT |ATT |MCI |MCI |MCI |ATT |ATT |ATT |ATT

| | | | | | | | | |MCI |MCI

------------------------------------------------------------

------------------------------------------------------------

|Chicago | | |Amer|Amer|Amer|MFS |MFS |MFS | |

------------------------------------------------------------

|Date -> |0317|0324|0331|0407|0414|0421|0428|0505|0512|0519

------------------------------------------------------------

After 05/05/97, the Chicago Primary and Tarrytown Backup would

become live with the LSMSs and SOAs that are hooked up and only

live data would be allowed after this. Lockheed says that 5/5 is a

clean-out date for all development testing data. The Chicago

test bed would then be used for all other testers. They are

also working on an Application simulator to allow vendors to

perform early turn up type testing prior to giving to their

customers.

As of 6/15, Ralph says all the LSMSs should wipe out all LNP

data and start with a clean slate at that time. No objections

offered. This might not be a problem if only live data was

available in the NPAC SMS after 05/05/97. We will work this

with the NPAC team once this schedule is officially

3. Other NPAC issues

i. NPAC design change has been requested to mark NXX

combinations as not accessible by certain carriers

upon their request. This will delay delivery by a week.

Ralph is concerned about depending on that feature to

get clean operation. It is crucial to wipe the slate

clean on 6/15 independent of whether that feature is

implemented or not.

Lockheed stated again that 5/5 is a clean-out date

for all development testing data. Only live data will

be in the NPAC SMS after that.

ii. It was brought up that the vendors that then came on

would have to sync up to the live data and how would that

be done?

NPAC keeps extensive time-stamped logs on when events

occurred. Could generate reports to show how quickly

inputs are logged by SMSs. No info available on how the

information is distributed to the network by the SMS.

iii. Question: how long to reload an SCP pair after an outage?

Suppose SCP pair and LSMS all lose data. Minimum 5.2 CMIP

transactions per second throughput. Also supports block

transactions.

Could deliver on tape for very large databases. Might

take 2-4 hours to dump tape, plus transport time, plus

load time. Could deliver same stuff by FTP.

LSMS and NPAC reliability are being considered as one

whole piece.

iv. Test team requested that any schedule problems be

socialized as soon as they are known. This is essential.

Lockheed agrees.

VI. Miscellaneous issues

1. Walt asked that every test plan writer reference the

appropriate figures in Appendix A.

2. Barry asked if anybody has any feedback on the choice

of Dick Dowd as project coordinator. Offered to take

the feedback privately.

The coordinator will be integrated into the team, and

results will appear on the ported.com web page.

Details of the job will be defined by the test team

and the operations team.

No one had a problem with Dick Dowd as the coordinator.

Barry will take that back to the steering committee.

3. Need to get a list of the NXXs and DNs for each company.

Send them to Walt prior to January meeting. Minimum number

of DNs per company has not been decided.

Crucially important to define clearly which NXXs and DNs

will be used, and to assure that only agreed-upon NXXs and

DNs are in the setup.

4. Issues list - Walt will gather up all of the issues and put them together in

a list to track for the January meeting. He will send out to everyone before

that meeting.

5. Schedule:

All updates due to Bill Belshaw by 12/24/96

He will have the updated issue 1 test plan on

www.cameos.com by Jan. 3

Barry will put up a zipped version of the files

on www.ported.com soon after that.

6. Next Meeting:

Jan 08 & 09

08:30 to 17:00

350 N. Orleans, Chicago

Room 437.

7. Conf. Call 01/27/96 at 09:00 on 312-814-8097