Research:
Deliverables: D5.1 Training and Dissemination Activities
Participants
The 2nd DESIRE II, Web Cache Managers Workshop was attended
by 19 participants coming from eight different European countries.
The representation by country was the following:
| Croatia |
1 |
| Cyprus |
1 |
| Hungary |
7 |
| Italy |
4 |
| Poland |
3 |
| Portugal |
1 |
| Spain |
1 |
| The Netherlands |
1 |
The following table provides details of the participants
and, whenever available, information about their activity and interests
in the caching area.
| Csaba Koller |
Concorde Direct |
Hungary |
|
| Dergez Bertalan |
POTE Informatikai Kozpont |
Hungary |
|
| Drazen Gemic |
CARNet |
Croatia |
Cache service administrator for
CARNet domain |
| Geza Turchanyi |
MATAV |
Hungary |
My task is just the selection
of the best tools for IP service provision, web cache included |
| Henny Bekker |
SRFnet (DESIRE II Project Team) |
The Netherlands |
|
| Joao Pagaime |
FCCN |
Portugal |
Currently evaluating a webcaching
solution to place in front of our international lines. We did squid load
testing on BSD and Linux that might interest other people |
| Jozsef Kadlecsik |
Research Institute for Particle
and Nuclear Physics |
Hungary |
Administrator of the cache servers
at the KFKI campus |
| Marina Buzzi |
IAT CNR |
Italy |
Cache administrator |
| Massimo Ianigro |
CNR |
Italy |
administers the main proxy server
of the CNR - Area di Ricerca, which is also used by other institutes in
Bari (15 institutes) and others located somewhere in Italy. He has been
involved in the administration in one of the main national proxy server
of the CNR |
| Beatrix Toth |
MTA SZTAKI |
Hungary |
WWW Administrator |
| Paola Tentoni |
CILEA |
Italy |
|
| Pasztor, Gyorgy |
Computer Centre University of
Szeged |
Hungary |
Daily management of web caches.
(Check its run, upgrade software, etc.) |
| Piotr Goczal |
Technical University of Lodz |
Poland |
Administrator of w3cache.man.lodz.pl
(cache for MAN in Lodz) |
| Piotr Konstanty |
Nicholas Copernicus University |
Poland |
administrator of squid program,
regional w3cache server |
| Rodrigo Castro |
REDIRIS |
Spain |
Management, Develop and Consulting |
| Tomasz Jablonski |
Poznan Supercomputing & Networking
Center |
Poland |
I'm a network services admin in
PSNC, and I take care about our w3cache service. Our server is one of biggest
in POL-34 (polish research, national ATM network), and is part of our national
cache mesh. Because of it I'm interested in optimizing structure of cache
meshes and interoperability. |
| Toth-Abonyi, Mihaly |
Computer Centre University of
Szeged |
Hungary |
|
| Andrea Segato |
CSP s.c. a r.l. |
Italy |
We are beginners in this field,
we want to deepen our knowledge |
| Yiannos Pitas |
University of Cyprus |
Cyprus |
The University of Cyprus and the
Cyprus Academic and Research Network consider the web cache technology
a very essential and crucial part of any modern network backbone.
For this reason, we are very interested in this new and evolving technology
that will boost our backbone networks and thus the services we offer |
Feedback
All participants were requested to fill in a Workshop
Evaluation Form.
16 forms out of 19 were returned to TERENA, who collected
the results summarised below.
Aims and content of the workshop
-
1) Relevance, average score 4.56 (on a maximum of 5)
16 responses: 7 rated the workshop as 4 and 9 as 5 (excellent)
-
2) State of the art, average score 4.25 (on a maximum of
5)
16 responses: 1 rated the workshop as 3, 10 as 4 and
5 as 5 (excellent)
-
3) Met objectives, average score 4.31 (on a maximum of 5)
16 responses: 1 rated the workshop as 3, 9 as 4 and 6
as 5 (excellent)
-
4) Clarity, average score 4.13 (on a maximum of 5)
16 responses: 3 rated the workshop as 3, 8 as 4 and 5
as 5 (excellent)
-
5) Value to participants, average score 4.69 (on a maximum
of 5)
16 responses: 5 rated the workshop as 4 and 11 as 5 (excellent)
The workshop sessions
6) The most useful sessions were, in the order:
-
4 Evaluation of caching products, 7 preferences
-
5 SQUID and MRTG, 6 preferences
-
6 Tutorial on cache meshes, 6 preferences
-
1 Cost/benefit and Caching architecture, 3 preferences
-
8 Development updates and standards, 3 preferences
-
2 Tutorial on SQUID, 2 preferences
-
3 Log analysis, 2 preferences
7) The most useless were, in the order:
-
1 Cost/benefit and Caching architecture, 3 preferences
-
7 transparent caching and FTP mirror tracking, 3 preferences
-
2 Tutorial on SQUID, 2 preferences
-
4 Evaluation of caching products, 2 preferences
-
6 Tutorial on cache meshes, 1 preference
-
8 Development updates and standards, 1 preference
8) Were there any topics that you think the workshop could
have covered, which were not covered?
I`d like more details on HTTP/1.1 protocol, cache directives
and protocols (WCCP, NECP).
Authentication problems in transparent caching and possible
solutions.
Proxy autoconfiguration issues.
[session N.] 8
[session N.] 8
No.
There are any way to transparent proxy ftp sessions.
(not netscape, …) ** think *** *** *** . Why, why not?
Web cache administration cost -setup time, keeping a
production installation
A relation of caching servers to network equipment
(load balancing, etc.)
Real-world scenarios.
The little details that we could do to fine tune the
caches (how many disks, how much we should fill them with objects, etc.)
Content filtering.
Facilities and organisation
The facilities and organisation got the following scores:
-
9) Pre-conference organisation, average score 3.87 (on a
maximum of 5)
15 responses: 1 rated it as 1(poor), 1 as 2, 1 as 3,
8 as 4 and 4 as 5 (excellent)
-
10) Conference organisation, average score 4.75 (on a maximum
of 5)
16 responses: 4 rated it as 4 and 12 as 5 (excellent)
-
11) Workshop material, average score 3.73 (on a maximum of
5)
15 responses: 1 rated it as 2, 3 as 3, 10 as 4 and 1
as 5 (excellent)
-
12) Workshop duration, average score 2.31 (on a maximum of
3)
Out of 16, 5 participants thought that the workshop was
too short, 11 thought that it was about right.
Relevance to participants
The participants answers about this set of open question
were very different in nature. Some words in a few sentences were particularly
difficult to understand, in those cases a number of stars "***" is used
to indicate the editors inability to decode the handwriting.
13) What I liked most about the whole workshop:
Both workshop contents and the possibility to exchange
ideas with other people!
The exchange of ideas and results between different managers.
The opportunity to meet people and hear about evolutions
in www caching.
Easy way of getting contact with participants or speakers
(small group of workshop members)
The professional level of most of the presentations
If someone doesn`t heard about anything before the WS,
he could *** the basics (but not ** order ? e.g. we heard about installing
SQUID before we ***, ** to organise servers)
Exchanging knowledge.
High technical standard of the classroom. Practical exercises
were prepared very well.
*** *** good overview
S.1.4
Reference to SQUID, cache architecture, future of caching
technology.
To meet people who are deeply involved in the topics
that were covered.
14) What I liked least about the whole workshop:
Practical sessions.
The cofee :-)
Some boring noise in the workshop room.
Practical session too short: free time
There were no information on what is covered by TERENA
and what should be paid by participants. Many slides were missing from
the workbook.
We didn`t know before the workshop things punctually.
E.g.: I didn`t know: we have to organise ourself the lunch or we get (but
fortunately we got :-))
Not enough time to configure SQID/MRTG with SNMP properly.
There was too warm in the room
No tea a place ;-)
Some of the hands-on (like compiling stuff for Linux
that everyone knows it will not fail!)
Little more things about the performance issues.
15) How could we have improved this workshop for you:
It was very good!
Download much of the files necessary to the practical
session before and also compile them. Let us just the configuration.
Better pre-conference information, more up-to-date workbook.
More "documented": also the sessions, & also the
event (e.g. ***)
Having more commercial products
Examine-reference real-world scenarios, problems encountered,
things to avoid/look for.
I will use them when configuring my cache in the future.
16) Did the workshop met your expectations,
There were 15 responses: 15 yes
17) Do you plan to use any of the information you have
been given at this workshop?
Yes. To organise a seminar for cache (& system/network)
administrator in our organisation.
- To evaluate my installation of caches.
- To improve the statistics of my caches.
- To make a real evaluation of cost/benefit.
Yes.
1. Upgrade to new version of SQUID.
2. Configure SQUID with new options suitable for my environment.
3. Use MRTG to get ** of *** SQUID, with SNMP
Info from 6 session.
SQUID, Cache system architecture.
Better quality reports about caches to the management.
A couple of ideas will be tried, tested in our environment.
Yes. I want to setup MRTG to tack my caches.
Not directly. Currently SQUID isn`t an option for our
1st level web cache system.
I am going to use it to make decisions about starting
cache servers in small institutions.
Yes. Nearly all.
Yes, definitely yes.
Yes!
18) Further comments and suggestions
Don`t use dark slide, in order to allow the "user" to
write on them.
Please put all showed slides into a downloadable ***.
Then I can view again all working example, and I think then I can try out
those things are easily
For real practical work, much *** would needed. However,
*** *** that practical exercises really needed.
|
