Hi everybody,

Let's suppose that our system is composed by a MTC as controller of several PTCs, and each PTC (declared as the same type of component) are running on different systems (although systems are similar, same number of ports, declared with the same name). Can we handle this architecture directly from TTCN-3 code? or if it is not possible, what kind of changes may it supposed on the SA?



To situation is as follow:







Thanks in advance,

Best Regards.



Sergio Lobato.

Hi Sergio,

if you mean with "running on different systems" real different physical
test devices, I would not see a problem. Because you can map ports from
different PTCs on the same port of the TSI.

your test system interface looks like

type component TSI {
port P p1, p2, p3 ;
}

each ptc can now do something like this:


function ptcF1() runs on PTC {
map(self:p, system:p1) ;
}

or

function ptcF2() runs on PTC {
map(self:p1, system:p1);
map(self:p2, system:p2);
}

with

type component PTC {
port P p1, p2, p3 ;
}

Typically the problem is that if the PTCs run on a single test device,
you have to multiplex/demultiplex your logical ptc-system mappings with
the real communication.

If you can guarantee that the PTCs are running on different devices, a
SA is typically simpler to implement, as you will see only on triMap per
system port.

Does this answer your question?

BR, Theo


Sergio Lobato López schrieb:
> Hi everybody,
>
> Let’s suppose that our system is composed by a MTC as controller of
> several PTCs, and each PTC (declared as the same type of component) are
> running on different systems (although systems are similar, same number
> of ports, declared with the same name). Can we handle this architecture
> directly from TTCN-3 code? or if it is not possible, what kind of
> changes may it supposed on the SA?
>
>
>
> To situation is as follow:
>
>
>
>
>
> Thanks in advance,
>
> Best Regards.
>
>
>
> Sergio Lobato.
>

--

---

Theofanis Vassiliou-Gioles email: vassiliou -at- testingtech.de
Testing Technologies IST GmbH phone: +49 30 726 19 190
Rosenthaler Str. 13 fax : +49 30 726 19 1920
D-10119 Berlin, Germany web : www.testingtech.de

Hi Theo,
It wasn't what I wanted to express, my concern comes up when I want to work with two different TSIs, one per PTC. Is this possible?

Thank you,
Regards.
Sergio

---

Mensaje original

---

De: active_ttcn3 : mts stf133 ttcn version 3 - active members only [This email address is being protected from spambots. You need JavaScript enabled to view it.] En nombre de Theofanis Vassiliou-Gioles
Enviado el: viernes, 23 de febrero de 2007 12:31
Para: This email address is being protected from spambots. You need JavaScript enabled to view it.
Asunto: Re: Several systems

Hi Sergio,

if you mean with "running on different systems" real different physical
test devices, I would not see a problem. Because you can map ports from
different PTCs on the same port of the TSI.

your test system interface looks like

type component TSI {
port P p1, p2, p3 ;
}

each ptc can now do something like this:


function ptcF1() runs on PTC {
map(self:p, system:p1) ;
}

or

function ptcF2() runs on PTC {
map(self:p1, system:p1);
map(self:p2, system:p2);
}

with

type component PTC {
port P p1, p2, p3 ;
}

Typically the problem is that if the PTCs run on a single test device,
you have to multiplex/demultiplex your logical ptc-system mappings with
the real communication.

If you can guarantee that the PTCs are running on different devices, a
SA is typically simpler to implement, as you will see only on triMap per
system port.

Does this answer your question?

BR, Theo


Sergio Lobato López schrieb:
> Hi everybody,
>
> Let's suppose that our system is composed by a MTC as controller of
> several PTCs, and each PTC (declared as the same type of component) are
> running on different systems (although systems are similar, same number
> of ports, declared with the same name). Can we handle this architecture
> directly from TTCN-3 code? or if it is not possible, what kind of
> changes may it supposed on the SA?
>
>
>
> To situation is as follow:
>
>
>
>
>
> Thanks in advance,
>
> Best Regards.
>
>
>
> Sergio Lobato.
>

--

---

Theofanis Vassiliou-Gioles email: vassiliou -at- testingtech.de
Testing Technologies IST GmbH phone: +49 30 726 19 190
Rosenthaler Str. 13 fax : +49 30 726 19 1920
D-10119 Berlin, Germany web : www.testingtech.de

Hello Sergio,

If one is just looking at the language, the answer is a simple 'no'. There is one system component (=TSI) in a test case. This is fine from the point of view of the TTCN-3 behaviour, where it does not matter on which hosts a PTC is executed and which test system port is related to whatever external interface of the SUT.

If you need some kind of distribution, this has to be captured in the implemenation of the SA. You can support this by mapping several PTC ports to one port at the TSI and use the component reference to resolve this. Alternatively you can also use port arrays at the test system interface.

From my own experience such distribution problems could be handled mostly by the SA, at least as long as functional testing is involved. In performance testing, the performance of the test system itself might become the limiting factor itself, in such test cases the test system itself might have to be distributed and then such a distributed TSI might be interesting. (I assume that Theo could tell us more about such distributed systems.)

Best regards

Thomas

---

| Thomas Deiß, Nokia Research Center Street address: |
| P.O. Box 101823 Meesmannstrasse 103 |
| D-44718 Bochum, GERMANY D-44807 Bochum, GERMANY |
| Phone: +49 234 984 2217, Mobile: +49 163 984 2217 |
| Fax: +49 234 984 3491, E-mail: This email address is being protected from spambots. You need JavaScript enabled to view it. |

---

>

---

Original Message

---

>From: active_ttcn3 : mts stf133 ttcn version 3 - active
>members only [This email address is being protected from spambots. You need JavaScript enabled to view it.] On Behalf Of ext
>Sergio Lobato López
>Sent: Friday, 23. February 2007 13:08
>To: This email address is being protected from spambots. You need JavaScript enabled to view it.
>Subject: Re: Several systems
>
>Hi Theo,
>It wasn't what I wanted to express, my concern comes up when I
>want to work with two different TSIs, one per PTC. Is this possible?
>
>Thank you,
>Regards.
>Sergio
>
>

---

Mensaje original

---

>De: active_ttcn3 : mts stf133 ttcn version 3 - active members
>only [This email address is being protected from spambots. You need JavaScript enabled to view it.] En nombre de Theofanis
>Vassiliou-Gioles Enviado el: viernes, 23 de febrero de 2007 12:31
>Para: This email address is being protected from spambots. You need JavaScript enabled to view it.
>Asunto: Re: Several systems
>
>Hi Sergio,
>
>if you mean with "running on different systems" real different
>physical test devices, I would not see a problem. Because you
>can map ports from different PTCs on the same port of the TSI.
>
>your test system interface looks like
>
>type component TSI {
> port P p1, p2, p3 ;
>}
>
>each ptc can now do something like this:
>
>
>function ptcF1() runs on PTC {
> map(self:p, system:p1) ;
>}
>
>or
>
>function ptcF2() runs on PTC {
> map(self:p1, system:p1);
> map(self:p2, system:p2);
>}
>
>with
>
>type component PTC {
> port P p1, p2, p3 ;
>}
>
>Typically the problem is that if the PTCs run on a single test
>device, you have to multiplex/demultiplex your logical
>ptc-system mappings with the real communication.
>
>If you can guarantee that the PTCs are running on different
>devices, a SA is typically simpler to implement, as you will
>see only on triMap per system port.
>
>Does this answer your question?
>
>BR, Theo
>
>
>Sergio Lobato López schrieb:
>> Hi everybody,
>>
>> Let's suppose that our system is composed by a MTC as controller of
>> several PTCs, and each PTC (declared as the same type of component)
>> are running on different systems (although systems are similar, same
>> number of ports, declared with the same name). Can we handle this
>> architecture directly from TTCN-3 code? or if it is not
>possible, what
>> kind of changes may it supposed on the SA?
>>
>>
>>
>> To situation is as follow:
>>
>>
>>
>>
>>
>> Thanks in advance,
>>
>> Best Regards.
>>
>>
>>
>> Sergio Lobato.
>>
>
>--
>

---

>Theofanis Vassiliou-Gioles email: vassiliou -at- testingtech.de
>Testing Technologies IST GmbH phone: +49 30 726 19 190
>Rosenthaler Str. 13 fax : +49 30 726 19 1920
>D-10119 Berlin, Germany web : www.testingtech.de
>

Hello,

Thomas is totally right. There is no way to have multiple TSI for one test case.
Usally you would need then at TSI which is the sum of all your single TSI.

And yes, if there are some more questions on "distributed testing" just let me
know.

Best regards,
Theo



Zitat von Thomas Deiss <This email address is being protected from spambots. You need JavaScript enabled to view it.>:

> Hello Sergio,
>
> If one is just looking at the language, the answer is a simple 'no'. There is
> one system component (=TSI) in a test case. This is fine from the point of
> view of the TTCN-3 behaviour, where it does not matter on which hosts a PTC
> is executed and which test system port is related to whatever external
> interface of the SUT.
>
> If you need some kind of distribution, this has to be captured in the
> implemenation of the SA. You can support this by mapping several PTC ports to
> one port at the TSI and use the component reference to resolve this.
> Alternatively you can also use port arrays at the test system interface.
>
> From my own experience such distribution problems could be handled mostly by
> the SA, at least as long as functional testing is involved. In performance
> testing, the performance of the test system itself might become the limiting
> factor itself, in such test cases the test system itself might have to be
> distributed and then such a distributed TSI might be interesting. (I assume
> that Theo could tell us more about such distributed systems.)
>
> Best regards
>
> Thomas
>
>

---

> | Thomas Deiß, Nokia Research Center Street address: |
> | P.O. Box 101823 Meesmannstrasse 103 |
> | D-44718 Bochum, GERMANY D-44807 Bochum, GERMANY |
> | Phone: +49 234 984 2217, Mobile: +49 163 984 2217 |
> | Fax: +49 234 984 3491, E-mail: This email address is being protected from spambots. You need JavaScript enabled to view it. |
>

---

>
>
>
> >

---

Original Message

---

> >From: active_ttcn3 : mts stf133 ttcn version 3 - active
> >members only [This email address is being protected from spambots. You need JavaScript enabled to view it.] On Behalf Of ext
> >Sergio Lobato López
> >Sent: Friday, 23. February 2007 13:08
> >To: This email address is being protected from spambots. You need JavaScript enabled to view it.
> >Subject: Re: Several systems
> >
> >Hi Theo,
> >It wasn't what I wanted to express, my concern comes up when I
> >want to work with two different TSIs, one per PTC. Is this possible?
> >
> >Thank you,
> >Regards.
> >Sergio
> >
> >

---

Mensaje original

---

> >De: active_ttcn3 : mts stf133 ttcn version 3 - active members
> >only [This email address is being protected from spambots. You need JavaScript enabled to view it.] En nombre de Theofanis
> >Vassiliou-Gioles Enviado el: viernes, 23 de febrero de 2007 12:31
> >Para: This email address is being protected from spambots. You need JavaScript enabled to view it.
> >Asunto: Re: Several systems
> >
> >Hi Sergio,
> >
> >if you mean with "running on different systems" real different
> >physical test devices, I would not see a problem. Because you
> >can map ports from different PTCs on the same port of the TSI.
> >
> >your test system interface looks like
> >
> >type component TSI {
> > port P p1, p2, p3 ;
> >}
> >
> >each ptc can now do something like this:
> >
> >
> >function ptcF1() runs on PTC {
> > map(self:p, system:p1) ;
> >}
> >
> >or
> >
> >function ptcF2() runs on PTC {
> > map(self:p1, system:p1);
> > map(self:p2, system:p2);
> >}
> >
> >with
> >
> >type component PTC {
> > port P p1, p2, p3 ;
> >}
> >
> >Typically the problem is that if the PTCs run on a single test
> >device, you have to multiplex/demultiplex your logical
> >ptc-system mappings with the real communication.
> >
> >If you can guarantee that the PTCs are running on different
> >devices, a SA is typically simpler to implement, as you will
> >see only on triMap per system port.
> >
> >Does this answer your question?
> >
> >BR, Theo
> >
> >
> >Sergio Lobato López schrieb:
> >> Hi everybody,
> >>
> >> Let's suppose that our system is composed by a MTC as controller of
> >> several PTCs, and each PTC (declared as the same type of component)
> >> are running on different systems (although systems are similar, same
> >> number of ports, declared with the same name). Can we handle this
> >> architecture directly from TTCN-3 code? or if it is not
> >possible, what
> >> kind of changes may it supposed on the SA?
> >>
> >>
> >>
> >> To situation is as follow:
> >>
> >>
> >>
> >>
> >>
> >> Thanks in advance,
> >>
> >> Best Regards.
> >>
> >>
> >>
> >> Sergio Lobato.
> >>
> >
> >--
> >

---

> >Theofanis Vassiliou-Gioles email: vassiliou -at- testingtech.de
> >Testing Technologies IST GmbH phone: +49 30 726 19 190
> >Rosenthaler Str. 13 fax : +49 30 726 19 1920
> >D-10119 Berlin, Germany web : www.testingtech.de
> >
>
>


--
Theofanis Vassiliou-Gioles email: This email address is being protected from spambots. You need JavaScript enabled to view it.
Testing Technologies IST GmbH phone: +49 30 726 19 190
Rosenthaler Str. 13 fax : +49 30 726 19 1920
D-10119 Berlin, Germany www.testingtech.de

Hello,
Thank you very much for your answers that have been really useful.

Regards,
Sergio.

---

Mensaje original

---

De: active_ttcn3 : mts stf133 ttcn version 3 - active members only [This email address is being protected from spambots. You need JavaScript enabled to view it.] En nombre de Theofanis Vassiliou-Gioles
Enviado el: viernes, 23 de febrero de 2007 16:03
Para: This email address is being protected from spambots. You need JavaScript enabled to view it.
Asunto: Re: Several systems

Hello,

Thomas is totally right. There is no way to have multiple TSI for one test case.
Usally you would need then at TSI which is the sum of all your single TSI.

And yes, if there are some more questions on "distributed testing" just let me
know.

Best regards,
Theo



Zitat von Thomas Deiss <This email address is being protected from spambots. You need JavaScript enabled to view it.>:

> Hello Sergio,
>
> If one is just looking at the language, the answer is a simple 'no'. There is
> one system component (=TSI) in a test case. This is fine from the point of
> view of the TTCN-3 behaviour, where it does not matter on which hosts a PTC
> is executed and which test system port is related to whatever external
> interface of the SUT.
>
> If you need some kind of distribution, this has to be captured in the
> implemenation of the SA. You can support this by mapping several PTC ports to
> one port at the TSI and use the component reference to resolve this.
> Alternatively you can also use port arrays at the test system interface.
>
> From my own experience such distribution problems could be handled mostly by
> the SA, at least as long as functional testing is involved. In performance
> testing, the performance of the test system itself might become the limiting
> factor itself, in such test cases the test system itself might have to be
> distributed and then such a distributed TSI might be interesting. (I assume
> that Theo could tell us more about such distributed systems.)
>
> Best regards
>
> Thomas
>
>

---

> | Thomas Deiß, Nokia Research Center Street address: |
> | P.O. Box 101823 Meesmannstrasse 103 |
> | D-44718 Bochum, GERMANY D-44807 Bochum, GERMANY |
> | Phone: +49 234 984 2217, Mobile: +49 163 984 2217 |
> | Fax: +49 234 984 3491, E-mail: This email address is being protected from spambots. You need JavaScript enabled to view it. |
>

---

>
>
>
> >

---

Original Message

---

> >From: active_ttcn3 : mts stf133 ttcn version 3 - active
> >members only [This email address is being protected from spambots. You need JavaScript enabled to view it.] On Behalf Of ext
> >Sergio Lobato López
> >Sent: Friday, 23. February 2007 13:08
> >To: This email address is being protected from spambots. You need JavaScript enabled to view it.
> >Subject: Re: Several systems
> >
> >Hi Theo,
> >It wasn't what I wanted to express, my concern comes up when I
> >want to work with two different TSIs, one per PTC. Is this possible?
> >
> >Thank you,
> >Regards.
> >Sergio
> >
> >

---

Mensaje original

---

> >De: active_ttcn3 : mts stf133 ttcn version 3 - active members
> >only [This email address is being protected from spambots. You need JavaScript enabled to view it.] En nombre de Theofanis
> >Vassiliou-Gioles Enviado el: viernes, 23 de febrero de 2007 12:31
> >Para: This email address is being protected from spambots. You need JavaScript enabled to view it.
> >Asunto: Re: Several systems
> >
> >Hi Sergio,
> >
> >if you mean with "running on different systems" real different
> >physical test devices, I would not see a problem. Because you
> >can map ports from different PTCs on the same port of the TSI.
> >
> >your test system interface looks like
> >
> >type component TSI {
> > port P p1, p2, p3 ;
> >}
> >
> >each ptc can now do something like this:
> >
> >
> >function ptcF1() runs on PTC {
> > map(self:p, system:p1) ;
> >}
> >
> >or
> >
> >function ptcF2() runs on PTC {
> > map(self:p1, system:p1);
> > map(self:p2, system:p2);
> >}
> >
> >with
> >
> >type component PTC {
> > port P p1, p2, p3 ;
> >}
> >
> >Typically the problem is that if the PTCs run on a single test
> >device, you have to multiplex/demultiplex your logical
> >ptc-system mappings with the real communication.
> >
> >If you can guarantee that the PTCs are running on different
> >devices, a SA is typically simpler to implement, as you will
> >see only on triMap per system port.
> >
> >Does this answer your question?
> >
> >BR, Theo
> >
> >
> >Sergio Lobato López schrieb:
> >> Hi everybody,
> >>
> >> Let's suppose that our system is composed by a MTC as controller of
> >> several PTCs, and each PTC (declared as the same type of component)
> >> are running on different systems (although systems are similar, same
> >> number of ports, declared with the same name). Can we handle this
> >> architecture directly from TTCN-3 code? or if it is not
> >possible, what
> >> kind of changes may it supposed on the SA?
> >>
> >>
> >>
> >> To situation is as follow:
> >>
> >>
> >>
> >>
> >>
> >> Thanks in advance,
> >>
> >> Best Regards.
> >>
> >>
> >>
> >> Sergio Lobato.
> >>
> >
> >--
> >

---

> >Theofanis Vassiliou-Gioles email: vassiliou -at- testingtech.de
> >Testing Technologies IST GmbH phone: +49 30 726 19 190
> >Rosenthaler Str. 13 fax : +49 30 726 19 1920
> >D-10119 Berlin, Germany web : www.testingtech.de
> >
>
>


--
Theofanis Vassiliou-Gioles email: This email address is being protected from spambots. You need JavaScript enabled to view it.
Testing Technologies IST GmbH phone: +49 30 726 19 190
Rosenthaler Str. 13 fax : +49 30 726 19 1920
D-10119 Berlin, Germany www.testingtech.de