SLAC WWW-Tech Mtg 8/6/97SLAC WWW-Tech Mtg 8/6/97
Attendees: Tony Johnson, Les Cottrell, Andrea Chan, John Halperin, Laurie Gennari, Joan Winters, Dennis Wisinski, Charlotte Hee, Ruth McDunn, Dennis Wisinski, Bebo White, George Crane.
Agenda:
A form has been set up at http://www-sldnt.slac.stanford.edu/web-server-register/, and one can see a list of servers that have applied for registration at: http://www-sldnt.slac.stanford.edu/web-server-register/list.asp. We agreed to recommend that Web servers on port 80 should have an alias starting with www-. There will be exceptions to this such as the Alta Vista forum server (forums.slac.stanford.edu). SLACVM:80 will be blocked from offsite.
We still need justifications for WWW and WWW-BFACTORY.
There are about 12 port 80 requests. We looked at further pruning since there will probably be growth in future. Tony agreed to get rid of SLDB2 and SLDB3. Questions were raised as to why SSRL needs 3 NT Web servers. There were a few requests for NT Web servers to access databases. Since Dennis will run a central NT Web server, and is not ready for production use, we agreed to provide interim offsite access until the central server is ready at which time it will be re-visited.
NGYMAC3 was requested to support Japanese, but it is unclear why the central Web server can't support this. It is a MacHTTP server. It is an 8080 port server so it's not super-critical, however we should understand the need since it may affect other things and future planning. PEPII-wienand1 has a Mac that could be run on the PEPII Mac server.
We discussed whether we are ready to "do it" on Monday. The consensus was that it would be good to do it Monday. We need to publicize via the SLAC Announcements, SLUO desk, the original email list of 90 people to whom the original requests for justifications were sent, the comp-out, the Help desk. The goal is to get this out on Thursday Aug 7, 1997 (the day after this meeting). In addition we will send a special email this ASAP to those people with requests for 8080 servers reminding them they may need to re-address them. We want to do this tonight since we need feedback soon to ensure we can go ahead on Monday. We will get Pat Kreitz to send out the Emails and be the lightening rod.
Charlotte will contact the folks for whom we need more information on their server justification. Tony will get together with Pat to put together the announcements, contact P. A. get posted etc. John will research the contacts for the most recent of newly discovered servers. If we encounter problems as a result of the blocking and the problems require a quick response, an emergency response team consisting of Tony, Pat, Charlotte, Dennis, John, Les have agreed to meet when needed.
The hardware can be retired but the URL needs to be preserved. Its URL is http://www6conf.slac.stanford.edu/ so it can't simply be put on the regular SLAC WWW server (e.g. with an alias) since it would then go to the SLAC home page rather than the www6conf home page. At the least, we will need a justification for it, Bebo will do this. Over time we will investigate moving it to another place (possibly offsite at W3C or the Web history site, or Stanford).
This looks like a very nice server (according to the glossies), it works well with the Oracle CASE tools. They anticipate it will become an "Enterprise Server" for the Oracle enterprise application. They will be testing for about a month. It would be good to use the default port if it is successful. They would like to not have to move to another port later on (e.g. from 8080 to 80). There are also bugs in Netscape that prevent it accessing hosts with name (including the ":" port-number if used) length exceeding 32 characters, this would give problems for a name like www-oracle.stanford.edu:8080. It was agreed to assign it to 80 and block it for a month until we get SLDB2, SLDB3 back, plus potential ones from SSRL.
Once a month, twice a month or whenever needed. It was agreed to reserve the 1st Wednesday of the month. If more meetings are needed they will be scheduled.