perfsonar-dev - Re: [pS-dev] Functional Testing the LS
Subject: perfsonar development work
List archive
- From: Maciej Glowiak <>
- To: Loukik Kudarimoti <>
- Cc: Michael Michalis <>, Jason Zurawski <>, Nicolas Simar <>, Ilias Tsompanidis <>,
- Subject: Re: [pS-dev] Functional Testing the LS
- Date: Wed, 04 Oct 2006 10:39:32 +0200
- Face: iVBORw0KGgoAAAANSUhEUgAAADAAAAAwAQMAAABtzGvEAAAABlBMVEUAAAD///+l2Z/dAAAA CXBIWXMAAEU1AABFNQF8gVf5AAAAB3RJTUUH1QYQDjo6uEWvwgAAAM5JREFUGNNN0LFqAkEUheGj KRZsfATrvENgYyH4APabxwgWGUUQC99BsNDCInUq7VImbbDZ0kayxBXMuN7jvTuKVh//mZlmQKZ1 EhQ8GAVgZECspEBdWQHRjR70KlgFKkoUaCw3ijSYQ4n5HfBK4a4jDcdDQPol/80Sr9BxZOOL4Fmr Jq8VBx7eopaSPvWGOm67fqol3j1q0XNs7Nk2cs6MU6gPNzf+ZGKQX4Ek8H6rAnFZnXB2vJxJcv8g C2P+WzL4tD+Txc4KydrIkh+eAdo01QbjQ84vAAAAAElFTkSuQmCC
- Organization: Poznan Supercomputing and Networking Center
Hi Loukik,
I went through the Michael's document again. I think it covers all the tests that should be done. One missing step, I can imagine, is LSCleanup linked with LSKeepalive.
Keepalives should be tested separately as an input request, but there is also LSCleanup, which is run by scheduler. I think functional tests should also cover such functionality.
LSCleanup removes not-up-to-date Lookup Information basing on special control information (timestamps). There are several cases to test.
Control information should be also tested with LSRegister/Update, LSDeregister, LSKeepalive (timestamps should be modified)
I guess Michael's tests will cover such tests, but I didn't found explicit statements about it in the document.
Of course I can answer all Michael's farther question regarding Lookup Service.
So, my conclusion is that the document in its current state is fine, but some words about testing LSCleanup should be added.
Maciej
Loukik Kudarimoti wrote:
Hi Maciej,
Michael is working on Functional tests or black-box testing. As you might already know, this process will test the functionalities specifications of a given software. It will do so with with the help of good and bad requests, good and bad configurations, etc.
What Michael requires is a functional specification of LS i.e. all the functionalities that are implemented and that you think are ready for testing and subsequent usage in the real world. This kind of functional specifications are normally expected to be written by the developers but we don't have any so far ( Please note that I am not trying to blame anybody here. We simply don't have those documents so far and we need them). Michael has written up such a document (in the long run, we will want these documents to contain more information about APIs and so on).
Can you please confirm that this document (in its current state) lists out all the interfaces that are ready to be documented in a user document used in the real world?
Thanks,
Loukik.
Maciej Glowiak wrote:
Michael,
I think your list is quite complete. I did most of these tests in order to test a service during the implementation.
You could use ant targets you mentioned in previous mail and just change the content XML files (LSRegisterRequest.xml, LSDeregisterRequest.xml and so on)
Maciej
Michael Michalis wrote:
Hi,
Attached to this message is a document describing the functional tests, that I intend to implement against the LS.
I would like to here your comments and suggestions on them.
Regards,
Michalis Michael
--
--------------------------------------------------------------------
| Maciej Glowiak Network Research and Development ||
|
Poznan Supercomputing and Networking Center ||
| (+48 61) 858 2024 http://monstera.man.poznan.pl/ ||
====================================================================
- Re: [pS-dev] Functional Testing the LS, Loukik Kudarimoti, 10/03/2006
- Re: [pS-dev] Functional Testing the LS, Maciej Glowiak, 10/04/2006
Archive powered by MHonArc 2.6.16.