From rory at siwaveinc.com Mon May 7 16:58:56 2001 From: rory at siwaveinc.com (Rory McKenna) Date: Mon, 7 May 2001 16:58:56 -0700 (PDT) Subject: Steve Vargo's coral access Message-ID: To whom it may concern, I wrote once before on this issue.. Steve Vargo (username "svargo") is having trouble using coral remotely. Is there any username and password we can use as a test which we know works for remote coral usage ? e.g. username: testcoral password: remotetest This would allow us to figure out if we can connect when we have a valid remote username/password. The next time Steve is up at Stanford, he's going to check/change his remote coral password so we can try again, but I would like to do a test with an authenticated username/pass in any case. Thank you, Rory From bmurray at snf.stanford.edu Wed May 9 08:19:41 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Wed, 9 May 2001 08:19:41 -0700 (PDT) Subject: Reboot Message-ID: John and Mike, I saw Len's note and that rosen had been rebooted. All the Coral logs looked good. I noticed the last activity in the logs took place at about 1:45am this morning. Do we know what the problem was? Thanks, Bill From pmei at exch.hpl.hp.com Tue May 15 13:48:41 2001 From: pmei at exch.hpl.hp.com (Mei, Ping) Date: Tue, 15 May 2001 13:48:41 -0700 Subject: Remote Coral Login Message-ID: <140D21516EC2D3119EE7009027876644043ADF10@hplex1.hpl.hp.com> Dear Coral Helper, I am trying to login onto Remote Coral from my home computer (window98 and no fire wall). I have set my remote password according to your instruction. However, I am not able to get to the equipment window. Here is the details of what happened: First, I launched Remote Coral. There was a login dialog window popped out. I input my user name and password. It was then quite for 1 ~ 2 mins and popped out an error message box: "Unable to receive equipment information". I wonder if you may know what is the problem and how to fix it? Thanks, -Ping ================================================== Ping Mei Hewlett-Packard Laboratories (650) 857-5137 1501 Page Mill Road (650) 852-8948 fax MS 4U-12 pmei at hpl.hp.com Palo Alto, CA 94304 ================================================== From bmurray at snf.stanford.edu Wed May 16 11:34:05 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Wed, 16 May 2001 11:34:05 -0700 (PDT) Subject: Active vs Inactive Message-ID: John and Mike, I am completing the implementation of the new resource manager methods that Mike and I added to the resource idl for the new client. In the current implementation, I'm only returning active members, projects, and accounts to load the trees and for searches. I don't want to clutter the trees and search results with inactive objects. However, we will probably want to provide at least the staff with some mechanism to see and/or search for inactive objects. Perhaps, they will need to re-activate projects, accounts, or members? Thanks, Bill From bmurray at snf.stanford.edu Sat May 19 18:09:43 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Sat, 19 May 2001 18:09:43 -0700 (PDT) Subject: Resource Module Message-ID: Mike and John, I've been making some changes to the resource server. I'm now maintaining sorted lists of the login names, projects, and accounts. I believe I can maintain these more efficiently than the clients and in a thread-safe manner. So Mike won't have to sort these (unless we see performance problems on the server side). Since additions and changes of members, projects, and accounts is a relatively infrequent event, I don't believe that it is necessary to use the event service to notify clients of these kinds of changes. Too much overhead for too little gain. Since we aren't using events, the clients will need to add their new members, projects, and accounts to their trees. Of course, the server will maintain complete and current names in the cache. The problem I'm worrying about is this. Ciara adds a new member, and Lourdes adds a new project. Neither one will see the other's changes reflected in their trees until they stop and restart their resource clients. However, if they do a search, they will be able to get all the information on the recently added project or member. Any comments and/or suggestions? Aaron Partridge had some minor, but good, suggestions for layout of the "Equipment Actions" pull-down menu. He suggested grouping "Shutdown" with problems and comments instead of with enable and disable. He also suggested changing the order of shutdown/problem/comment to "Make Comment", "Report Problem", and finally "Shutdown" in order of least serious to most serious. These are easy to change while I'm trying to integrate the new resource client into our client framework. Would you like me to do that? Now, on to integrating the resource client template into our gui. Bill From mbell at snf.stanford.edu Mon May 21 07:20:26 2001 From: mbell at snf.stanford.edu (Mike Bell) Date: Mon, 21 May 2001 07:20:26 -0700 Subject: Resource Module References: Message-ID: <3B09242A.AECD9D8@snf.stanford.edu> Bill, Won't I still need to sort the names / logons because I really need them by last, first (asc) and by logon (asc) depending on which type of search is being done? Mike Bill Murray wrote: > Mike and John, > > I've been making some changes to the resource server. I'm now maintaining > sorted lists of the login names, projects, and accounts. I believe I can > maintain these more efficiently than the clients and in a thread-safe > manner. So Mike won't have to sort these (unless we see performance > problems on the server side). > > Since additions and changes of members, projects, and accounts is a > relatively infrequent event, I don't believe that it is necessary to > use the event service to notify clients of these kinds of changes. Too > much overhead for too little gain. Since we aren't using events, the > clients will need to add their new members, projects, and accounts to their > trees. Of course, the server will maintain complete and current names > in the cache. The problem I'm worrying about is this. Ciara adds a new > member, and Lourdes adds a new project. Neither one will see the other's > changes reflected in their trees until they stop and restart their resource > clients. However, if they do a search, they will be able to get all the > information on the recently added project or member. Any comments and/or > suggestions? > > Aaron Partridge had some minor, but good, suggestions for layout of the > "Equipment Actions" pull-down menu. He suggested grouping "Shutdown" > with problems and comments instead of with enable and disable. He also > suggested changing the order of shutdown/problem/comment to > "Make Comment", "Report Problem", and finally "Shutdown" in order of > least serious to most serious. These are easy to change while I'm trying > to integrate the new resource client into our client framework. Would you > like me to do that? > > Now, on to integrating the resource client template into our gui. > > Bill > > From bmurray at snf.stanford.edu Mon May 21 16:03:39 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Mon, 21 May 2001 16:03:39 -0700 (PDT) Subject: MAILFORWARD / Coral (fwd) Message-ID: I watched Ciara set the forwarding address for hjcho. I believe she has been doing everything right. She expected a message telling her that the command had been successful. I told her you usually get a message when you do something wrong. Bill ---------- Forwarded message ---------- Date: Mon, 21 May 2001 14:48:55 -0700 From: Ciara Preston To: Bill Murray Subject: MAILFORWARD / Coral Bill, I've tried to change several peoples email, without success. I'm giving the command - Mailforward,then, when I get the prompt I type in mailforward again, a space, then their login, a space, next, quotation marks on both sides of their email address, then return. Is there something obvious that I'm doing wrong? Perchance could you come to my workspace & walk me through it? Ciara P. From bmurray at snf.stanford.edu Tue May 22 08:38:44 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Tue, 22 May 2001 08:38:44 -0700 (PDT) Subject: Status Message-ID: John and Mike, I encountered some difficulties/problems running the new resource client as a distinct executable, especially remotely. However, I believe I am close on a version that will meet all our requirements. I should have something to look at first thing tomorrow. Bill From pmei at exch.hpl.hp.com Tue May 22 09:24:17 2001 From: pmei at exch.hpl.hp.com (Mei, Ping) Date: Tue, 22 May 2001 09:24:17 -0700 Subject: Remote Coral Login Message-ID: <140D21516EC2D3119EE7009027876644043ADF21@hplex1.hpl.hp.com> I am still unable to remote login to Coral. I am using DSL line for my home computer. I wonder if Coral works with DSL connection. (It seems that Coral recognize my remote password. However, it does not bring the equipment window.) If you have a solution for this problem, please let me know. Thanks, -Ping -----Original Message----- From: Mei, Ping Sent: Tuesday, May 15, 2001 1:49 PM To: 'coral at snf.stanford.edu' Subject: Remote Coral Login Dear Coral Helper, I am trying to login onto Remote Coral from my home computer (window98 and no fire wall). I have set my remote password according to your instruction. However, I am not able to get to the equipment window. Here is the details of what happened: First, I launched Remote Coral. There was a login dialog window popped out. I input my user name and password. It was then quite for 1 ~ 2 mins and popped out an error message box: "Unable to receive equipment information". I wonder if you may know what is the problem and how to fix it? Thanks, -Ping ================================================== Ping Mei Hewlett-Packard Laboratories (650) 857-5137 1501 Page Mill Road (650) 852-8948 fax MS 4U-12 pmei at hpl.hp.com Palo Alto, CA 94304 ================================================== From bmurray at snf.stanford.edu Tue May 22 09:28:30 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Tue, 22 May 2001 09:28:30 -0700 (PDT) Subject: Resource Client & Security Message-ID: John and Mike, We should also discuss what level of security we need to provide in the resource client. For example, adding a new member from the remote version of Coral's resource client may introduce new risks. Bill From bmurray at snf.stanford.edu Tue May 22 12:56:30 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Tue, 22 May 2001 12:56:30 -0700 (PDT) Subject: Your Suggestion Message-ID: Aaron, I spoke with John regarding your suggestion for improvements to the "Equipment Actions" menu. He agreed with your suggestion to group "shutdown" with "report problem" and "make comment". However, he wanted to keep them in the same order (shutdown first, comment last). I've made the first change, grouping the three together, in the development version, and it will be released when we make the new resource client available next month. We really appreciate all your thoughtful suggestions. I also wanted to let you know that we haven't forgotten about your suggestion to change "enable" of a currently-enabled machine to begin charging the new user without physically disabling the machine. This improvement is still on our list. I hope to get to it soon. (Of course, we will warn the member that the machine is currently enabled by someone else before proceeding.) Thanks for your patience, Bill From beckwith at snf.stanford.edu Tue May 22 18:51:47 2001 From: beckwith at snf.stanford.edu (Sharleen Beckwith) Date: Tue, 22 May 2001 18:51:47 -0700 (PDT) Subject: lost netscape... Message-ID: how do I get it back? Thanks. From bmurray at snf.stanford.edu Wed May 23 08:38:54 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Wed, 23 May 2001 08:38:54 -0700 (PDT) Subject: Status Message-ID: John and Mike, I'm in the office and coding furiously. It's going great. I'll let you know when I have something to look at. (We'll have this knocked out in no time.) Bill From bmurray at snf.stanford.edu Wed May 23 10:33:24 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Wed, 23 May 2001 10:33:24 -0700 (PDT) Subject: New Sunrays at Door Message-ID: John and Mike, Mahnaz just stopped by to let me know how nice the new Sunrays are. Apparently she and some other lab members believe they are so nice that they might just walk away. She suggests securing them to the table in some way. What do you think? Bill From beckwith at snf.stanford.edu Wed May 23 11:23:44 2001 From: beckwith at snf.stanford.edu (Sharleen Beckwith) Date: Wed, 23 May 2001 11:23:44 -0700 (PDT) Subject: amtetcher history Message-ID: Could someone look in the history for last night and tell me who the last person was who used the amt? It doesn't appear on my coral history window. Thanks. From beckwith at snf.stanford.edu Wed May 23 11:30:42 2001 From: beckwith at snf.stanford.edu (Sharleen Beckwith) Date: Wed, 23 May 2001 11:30:42 -0700 (PDT) Subject: amtetcher history In-Reply-To: Message-ID: OK. I logged out and logged back in (those "smart" cards are not so smart after all, I guess they don't update). I now see the history for yesterday. mbadi said the machine was fine when he left it last night. Is there any way to tell who used it this morning before me? On Wed, 23 May 2001, Sharleen Beckwith wrote: > Could someone look in the history for last night and tell me who > the last person was who used the amt? It doesn't appear on my > coral history window. > > Thanks. > > From bmurray at snf.stanford.edu Wed May 23 11:45:50 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Wed, 23 May 2001 11:45:50 -0700 (PDT) Subject: amtetcher history In-Reply-To: Message-ID: Sharleen, These are the last three uses of the amtetcher. dtchang for 65 minutes from 2001-05-22 17:40:18 to 2001-05-22 18:44:34 mbadi for 55 minutes from 2001-05-22 23:16:36 to 2001-05-23 00:10:54 mbadi for 74 minutes from 2001-05-23 10:05:28 to 2001-05-23 11:18:46 According to the logs: dtchang enabled and disable the first run mbadi enabled and disbled the second run mbadi enabled and beckwith disabled the third run. Bill On Wed, 23 May 2001, Sharleen Beckwith wrote: > OK. I logged out and logged back in (those "smart" cards are not > so smart after all, I guess they don't update). I now see the > history for yesterday. mbadi said the machine was fine when he > left it last night. Is there any way to tell who used it this > morning before me? > > On Wed, 23 May 2001, Sharleen Beckwith wrote: > > > Could someone look in the history for last night and tell me who > > the last person was who used the amt? It doesn't appear on my > > coral history window. > > > > Thanks. > > > > > From mtang at snf.stanford.edu Wed May 23 16:53:17 2001 From: mtang at snf.stanford.edu (Mary Tang) Date: Wed, 23 May 2001 16:53:17 -0700 Subject: Login name... Message-ID: <3B0C4D6C.E96C7974@snf.stanford.edu> Greetings Coral Development Team -- Help. I'm trying to qualify Michele Cash (login name "michelec") on the wafersaw, but Coral doesn't seem to recognize this login name. "Finger" on SNF, however, does show she has an SNF account under this name. (Three other qualifications went fine today...) Is there a problem with her account? Thanks, Mary -- Mary X. Tang, Ph.D. National Nanofabrication Users' Network Stanford Nanofabrication Facility CIS Room 136, Mail Code 4070 Stanford, CA 94305 (650)723-9980 mtang at snf.stanford.edu From bmurray at snf.stanford.edu Wed May 23 17:05:52 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Wed, 23 May 2001 17:05:52 -0700 (PDT) Subject: Login name... In-Reply-To: <3B0C4D6C.E96C7974@snf.stanford.edu> Message-ID: Mary, Michele Cash's login name was entered in the Coral database as mcash. However, her account on snf and the sunray have been set up under the name michelec. I don't see how she can even get access to Coral under these circumstances. Bill On Wed, 23 May 2001, Mary Tang wrote: > Greetings Coral Development Team -- > > Help. > > I'm trying to qualify Michele Cash (login name "michelec") on the > wafersaw, but Coral doesn't seem to recognize this login name. "Finger" > on SNF, however, does show she has an SNF account under this name. > (Three other qualifications went fine today...) Is there a problem with > her account? > > Thanks, > > Mary > > -- > Mary X. Tang, Ph.D. > National Nanofabrication Users' Network > Stanford Nanofabrication Facility > CIS Room 136, Mail Code 4070 > Stanford, CA 94305 > (650)723-9980 > mtang at snf.stanford.edu > > From shott at snf.stanford.edu Wed May 23 17:22:23 2001 From: shott at snf.stanford.edu (John Shott) Date: Wed, 23 May 2001 17:22:23 -0700 Subject: Login name... References: Message-ID: <3B0C543F.D71F4F23@snf.stanford.edu> Mary: At the moment, the name that coral knows a person by and the name that they log in as are two separate names ... although they are supposed to be the same and, as Bill points out, Coral won't work properly for them if they are different. So, I've changed the database so that Michele's coral name matches her login name: michelec. You should now be able to qualify her ... Thanks, John From shott at snf.stanford.edu Wed May 23 17:32:24 2001 From: shott at snf.stanford.edu (John Shott) Date: Wed, 23 May 2001 17:32:24 -0700 Subject: Installation of software .... Message-ID: <3B0C5698.72AE0BAD@snf.stanford.edu> Bill and Mike: I'm in the process of trying to clean up some of our installation of software. In particular, I'm putting the sources in /app/solaris/2.6 ... on both guilden and sunray because they are, at the moment, distinct. Actually, since we have eliminated the development network, we can probably have a single /app that is mounted across all machines. In general, I'm doing make installs so that things end up in their default location ... usually /usr/local/bin. I'm also beginning to do some cleanup ... for example, we used to have bash and make in /app/tool/bin. I've removed those so that you should now automatically get the ones in /usr/local/bin. (Actually, I should check our Makefiles for labnet to make sure that we weren't explicitly calling /app/tool/bin/make ....). In any event, I hope that this will begin to clean things up for us and make it easier to tell penn state what they need to install. If I do a "ls -l" in /app/solaris/2.6 they should see the versions of everything that we are running. Does this sound reasonable? Also, does anyone care if I begin to blow away source files of old versions of things like ssh, bash, xemacs, and emacs (is everyone now using either /usr/local/bin/emacs or /usr/local/bin/xemacs?) Thanks, John From mtang at snf.stanford.edu Wed May 23 18:04:56 2001 From: mtang at snf.stanford.edu (Mary Tang) Date: Wed, 23 May 2001 18:04:56 -0700 Subject: Login name... References: <3B0C543F.D71F4F23@snf.stanford.edu> Message-ID: <3B0C5E38.DE06F425@snf.stanford.edu> Thanks!! I've also just emailed her to let her know. Mary John Shott wrote: > Mary: > > At the moment, the name that coral knows a person by and the name that they > log in as are two separate names ... although they are supposed to be the same > and, as Bill points out, Coral won't work properly for them if they are > different. So, I've changed the database so that Michele's coral name matches > her login name: michelec. > > You should now be able to qualify her ... > > Thanks, > > John -- Mary X. Tang, Ph.D. National Nanofabrication Users' Network Stanford Nanofabrication Facility CIS Room 136, Mail Code 4070 Stanford, CA 94305 (650)723-9980 mtang at snf.stanford.edu From anolley at iridigm.com Thu May 24 10:45:31 2001 From: anolley at iridigm.com (Alan Nolley, Sr.) Date: Thu, 24 May 2001 10:45:31 -0700 Subject: Help on Remote Coral access Message-ID: <3B216BAC294ED511984E00B0D079E00E02AF65@MAIL> Tech Support, We are trying to use the Remote CORAL and tried to get access to your equipment and we are getting an error message " Cannot Access equipment information ". The way we are setup in our network is we are behind a firewalland I was wondering if I need to open a port on our firewall to allow CORAL connection to happen? Or maybe there is some additional things we might need to configure in our end that you can tell us.... Any help will be appreciated. By the way this is for one of your registered user Alan Nolley. Joevie Bandalan Phone: 415 585 5481 Cell: 415 740 5373 joevie at systemsinnovation.com From mbell at snf.stanford.edu Thu May 24 11:11:36 2001 From: mbell at snf.stanford.edu (Mike Bell) Date: Thu, 24 May 2001 11:11:36 -0700 Subject: Help on Remote Coral access References: <3B216BAC294ED511984E00B0D079E00E02AF65@MAIL> Message-ID: <3B0D4ED7.7B7BF01A@snf.stanford.edu> Alan, I've talked to your systems person concerning the issues with Coral access. Our advice on this is to put a machine outside the firewall for Coral accesss. As part of my research on this problem I've identified a number of ports that might need to be opened up to allow Coral access. I mentioned this as well and suggested that if he wanted, Joevie could give this a try, although we can't be sure that the list I gave him is comprehensive and not subject to change. Please let me know if this helps. Thanks, Mike "Alan Nolley, Sr." wrote: > Tech Support, > We are trying to use the Remote CORAL and tried to get access to your > equipment and we are getting an error message " Cannot Access equipment > information ". The way we are setup in our network is we are behind a > firewalland I was wondering if I need to open a port on our firewall to > allow CORAL connection to happen? Or maybe there is some additional things > we might need to configure in our end that you can tell us.... > Any help will be appreciated. > By the way this is for one of your registered user Alan Nolley. > Joevie Bandalan > Phone: 415 585 5481 > Cell: 415 740 5373 > joevie at systemsinnovation.com From bmurray at snf.stanford.edu Thu May 24 18:58:53 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Thu, 24 May 2001 18:58:53 -0700 (PDT) Subject: Internet acces for Industrial users In-Reply-To: <3B0D88C4.B83623CF@snf.stanford.edu> Message-ID: Ciara, They need to contact an Internet Service Provider (ISP) like Pacific Bell. Bill On Thu, 24 May 2001, Ciara Preston wrote: > > Hi, Bill, > > Do you know how the Industrial users get an internet access line? > > Ciara > From shott at snf.stanford.edu Sun May 27 11:37:15 2001 From: shott at snf.stanford.edu (John Shott) Date: Sun, 27 May 2001 11:37:15 -0700 Subject: Coral References: <096C97EE8172D311AC19009027B11476012FEE7A@e-mail.engr.psu.edu> Message-ID: <3B11495B.8213151B@snf.stanford.edu> Mike: While Mike Bell and Bill Murray are working hard to finish the new-and-improved resource client (that will make it easy to add new lab members, projects, and accounts to the system ...), I thought that I'd try to send you some of the details of things that you will need to actually get, install, and run the coral system. Most of the "utility" things that we need are from GNU (www.gnu.org). Virtually all of their sofware if available for download at their site ... or, they have a number of "mirror" sites. For example, if you point your browser to ftp://boron.club.cc.cmu.edu/gnu/ you will get to what I think may be the closest mirror to you at Carnegie Mellon Univ. In general, installing gnu stuff is very easy ... generally, it will be downloaded as a "compressed tar file". Here is the procedure that I generally use: 1. Download the compressed tar file into your home directory ... in most cases it will be named something like: ~mrogosky/gnu_software_name-version.tar.gz where version is something like 2.3.0. 2. Go to the directory where you want to store the source files of this software. On our machine it is /app/solaris/2.6. 3. Become root (and make sure that you are still in your equivalent of /app/solaris/2.6. 4. Issue the command: tar -xzf ~rogosky/gnu_software_name-version.tar.gz This will uncompress and unpack the source files in a directory named gnu_software_name.verion 5. Enter that directory. 6. Configure, make and install that program. This usually involves issuing the commands: sh ./configure (Run the configure script) make (Which compiles everything) make install (Which installs the compiled code). We typically accept all of the defaults ... including the default installation location that is typically /usr/local/bin. Follwoign is the listing of software packages that we have on our machine in /app/solaris/2.6 (note: you don't have to use this name for the location of where you want to store stuff) .... Script started on Sun May 27 10:22:19 2001 [sunray:/app/solaris/2.6]$ pwd /app/solaris/2.6 [sunray:/app/solaris/2.6]$ ls -l total 68 -rw-r--r-- 1 root other 135 May 26 10:38 README -rw-r--r-- 1 root other 125 May 26 10:36 README~ drwxr-xr-x 2 1852 users 512 Jan 27 09:56 SSOLRS.install drwxr-xr-x 11 286 root 3584 May 26 12:12 bash-2.05 drwxr-xr-x 2 8482 staff 512 Jan 27 09:32 communicator-v476.sparc-sun-solaris2.5.1 drwxr-xr-x 14 501 501 1536 May 27 09:43 cvs-1.11.1p1 drwxr-xr-x 3 shott devel 1024 Jan 20 08:45 eieio-0.15 drwxr-xr-x 2 shott devel 1024 Jan 20 08:56 elib-1.0 drwxr-xr-x 16 gerd root 1024 Jan 18 16:59 emacs-20.7 drwxr-xr-x 7 root other 512 May 26 10:40 gcc drwxr-xr-x 16 4030 4030 1024 Mar 16 06:15 gcc-2.95.3 drwxr-xr-x 5 root root 1024 May 27 09:10 git-4.3.20 drwxr-xr-x 10 root root 1536 May 26 12:45 gzip-1.2.4a drwxr-xr-x 5 shott devel 512 Dec 21 21:17 jde-2.2.6 drwxr-xr-x 6 390 140 3072 May 23 16:46 make-3.79.1 drwxr-xr-x 2 shott devel 1024 Jan 20 08:47 semantic-1.3.2 drwxr-xr-x 3 shott devel 1024 Jan 20 08:48 speedbar-0.13a drwxr-xr-x 4 17275 20 4096 Jan 22 12:17 ssh-1.2.31 drwxr-xr-x 5 24814 users 1536 Jan 22 14:03 ssh-2.4.0 drwxr-xr-x 2 root root 512 May 23 15:27 stripcmt-0.1.2 drwxr-xr-x 11 1011 1011 1024 May 26 12:42 tar-1.13 drwxr-xr-x 10 bmurray root 1024 Jan 19 17:14 texinfo-4.0 drwxr-xr-x 14 501 501 1024 Jan 22 16:02 xemacs-21.1.13 script done on Sun May 27 10:36:12 2001 Here is what all of these things are and my recommendation as to the order in which to install them (grouped by functional areas ...). 1. The GNU C-compiler ... gcc. While your machine may already have a C-compiler, you probably want to make sure that you have a reasonably new version of gcc. You can test this by issuing the command "which gcc" and then "gcc -version". Above, you can see that we have gcc version 2.95.3. The compilation of this is a little non-standard ... and will take more time that virtually everything else. Sources are in /app/solaris/2.6 (or equivalent), then in that directory I would issue the following commands (I've annoated them with comments in () ... don't type that part. mkdir gcc (Make the "build" directory) cd gcc (Move into the build directory) sh ../gcc-2.95.3/configure (Run the configuration script in the source directory) make bootstrap (This will make the gcc compiler from the ground up) make install (This should install gcc in /usr/local/bin as well as install a bunch of libraries, header files, etc.) 2. Basic utilities: make, tar, gzip, and cvs. (Note: your machine probably already has a version of make, tar, and gzip ... and those may be just fine to use). I don't know to what degree these things will make a difference. The make, tar, and gzip utilities are available from the GNU mirror. CVS (Concurrent Versioning System) is avaialble from: cvshome.org. They all build with the standard "sh ./configure", "make", and "make install". Note: at this point it isn't clear whether we will initailly send you a compressed, tarred file for the Coral source or have you extract it directly from our CVS repository ... 3. Other useful things: bash, git, ssh. Bash (a shell) and git (for Gnu Interactive Tools) are both available from gnu. ssh (Secure Shell) is available from www.ssh.com. Ssh is sort of a secure version of ftp and telnet rolled into one. Around here, our computer security folks are increasingly reeluctant to allow us to use either telnet or ftp because both of those tools allow login names and passwords across the network "in the clear" ... anyone with a sniffer can easily learn lots of login names and passwords that way. Assuming that we will likely ultimately need accounts on your machine to help configure/troubleshoot your system, you should probalby have ssh installed. Note: we actually run both the older ssh1 as well as the newer ssh2 protocols because there are places on campus here that only run ssh1. If you can, you are better off getting and installing only ssh2 (we have version 2.4.0) ... both because it is easier to install and configure and because it is more secure. Oh yes, one other thing: stripcmt ... this is a simple program that strips comments out of C and java programs that we use. It is availabe from: 4. Editing environment for java source files. We use either emacs (available from GNU) or xemacs (available from www.xemacs.org) as our primary text editor. In addtion, we run a package called JDE (for Java Development Environment) on top of that. If you expect to be modifying source files ... and either you will or we will at your site ... it probalby makes sense to have a JDE/emacs or JDE/xemacs installation. JDE is availabe at www.sunsite.dk/jde You will see that this also requires the installation of several other packages including: elib, eieio, semantic and speedbar. This is actually the trickiest thing to install ... but the JDE web site gives pretty good instructions for installing it with either emacs or xemacs. Let me know if you have any problems getting this stuff installed and configured on your machine. Hopefully, having these tools and utilities available will make the "real installation" of Coral much less painless. If you find these instuctions and comments confusing, please let me know ... I'm hoping to use this as the "version 0" set of "HowTo Coral". Thanks, John From sfonash at psu.edu Sun May 27 17:48:02 2001 From: sfonash at psu.edu (Stephen J. Fonash) Date: Sun, 27 May 2001 20:48:02 -0400 Subject: Coral Message-ID: <4D99161EC716D411827B00B0D021EA2D373CBD@e-mail2.engr.psu.edu> John, Thanks for keeping this going. Steve Fonash -----Original Message----- From: John Shott [mailto:shott at snf.stanford.edu] Sent: Sunday, May 27, 2001 2:37 PM To: Michael Rogosky Cc: coral at snf.stanford.edu; Jeff Catchmark; Stephen J. Fonash Subject: Re: Coral Mike: While Mike Bell and Bill Murray are working hard to finish the new-and-improved resource client (that will make it easy to add new lab members, projects, and accounts to the system ...), I thought that I'd try to send you some of the details of things that you will need to actually get, install, and run the coral system. Most of the "utility" things that we need are from GNU (www.gnu.org). Virtually all of their sofware if available for download at their site ... or, they have a number of "mirror" sites. For example, if you point your browser to ftp://boron.club.cc.cmu.edu/gnu/ you will get to what I think may be the closest mirror to you at Carnegie Mellon Univ. In general, installing gnu stuff is very easy ... generally, it will be downloaded as a "compressed tar file". Here is the procedure that I generally use: 1. Download the compressed tar file into your home directory ... in most cases it will be named something like: ~mrogosky/gnu_software_name-version.tar.gz where version is something like 2.3.0. 2. Go to the directory where you want to store the source files of this software. On our machine it is /app/solaris/2.6. 3. Become root (and make sure that you are still in your equivalent of /app/solaris/2.6. 4. Issue the command: tar -xzf ~rogosky/gnu_software_name-version.tar.gz This will uncompress and unpack the source files in a directory named gnu_software_name.verion 5. Enter that directory. 6. Configure, make and install that program. This usually involves issuing the commands: sh ./configure (Run the configure script) make (Which compiles everything) make install (Which installs the compiled code). We typically accept all of the defaults ... including the default installation location that is typically /usr/local/bin. Follwoign is the listing of software packages that we have on our machine in /app/solaris/2.6 (note: you don't have to use this name for the location of where you want to store stuff) .... Script started on Sun May 27 10:22:19 2001 [sunray:/app/solaris/2.6]$ pwd /app/solaris/2.6 [sunray:/app/solaris/2.6]$ ls -l total 68 -rw-r--r-- 1 root other 135 May 26 10:38 README -rw-r--r-- 1 root other 125 May 26 10:36 README~ drwxr-xr-x 2 1852 users 512 Jan 27 09:56 SSOLRS.install drwxr-xr-x 11 286 root 3584 May 26 12:12 bash-2.05 drwxr-xr-x 2 8482 staff 512 Jan 27 09:32 communicator-v476.sparc-sun-solaris2.5.1 drwxr-xr-x 14 501 501 1536 May 27 09:43 cvs-1.11.1p1 drwxr-xr-x 3 shott devel 1024 Jan 20 08:45 eieio-0.15 drwxr-xr-x 2 shott devel 1024 Jan 20 08:56 elib-1.0 drwxr-xr-x 16 gerd root 1024 Jan 18 16:59 emacs-20.7 drwxr-xr-x 7 root other 512 May 26 10:40 gcc drwxr-xr-x 16 4030 4030 1024 Mar 16 06:15 gcc-2.95.3 drwxr-xr-x 5 root root 1024 May 27 09:10 git-4.3.20 drwxr-xr-x 10 root root 1536 May 26 12:45 gzip-1.2.4a drwxr-xr-x 5 shott devel 512 Dec 21 21:17 jde-2.2.6 drwxr-xr-x 6 390 140 3072 May 23 16:46 make-3.79.1 drwxr-xr-x 2 shott devel 1024 Jan 20 08:47 semantic-1.3.2 drwxr-xr-x 3 shott devel 1024 Jan 20 08:48 speedbar-0.13a drwxr-xr-x 4 17275 20 4096 Jan 22 12:17 ssh-1.2.31 drwxr-xr-x 5 24814 users 1536 Jan 22 14:03 ssh-2.4.0 drwxr-xr-x 2 root root 512 May 23 15:27 stripcmt-0.1.2 drwxr-xr-x 11 1011 1011 1024 May 26 12:42 tar-1.13 drwxr-xr-x 10 bmurray root 1024 Jan 19 17:14 texinfo-4.0 drwxr-xr-x 14 501 501 1024 Jan 22 16:02 xemacs-21.1.13 script done on Sun May 27 10:36:12 2001 Here is what all of these things are and my recommendation as to the order in which to install them (grouped by functional areas ...). 1. The GNU C-compiler ... gcc. While your machine may already have a C-compiler, you probably want to make sure that you have a reasonably new version of gcc. You can test this by issuing the command "which gcc" and then "gcc -version". Above, you can see that we have gcc version 2.95.3. The compilation of this is a little non-standard ... and will take more time that virtually everything else. Sources are in /app/solaris/2.6 (or equivalent), then in that directory I would issue the following commands (I've annoated them with comments in () ... don't type that part. mkdir gcc (Make the "build" directory) cd gcc (Move into the build directory) sh ../gcc-2.95.3/configure (Run the configuration script in the source directory) make bootstrap (This will make the gcc compiler from the ground up) make install (This should install gcc in /usr/local/bin as well as install a bunch of libraries, header files, etc.) 2. Basic utilities: make, tar, gzip, and cvs. (Note: your machine probably already has a version of make, tar, and gzip ... and those may be just fine to use). I don't know to what degree these things will make a difference. The make, tar, and gzip utilities are available from the GNU mirror. CVS (Concurrent Versioning System) is avaialble from: cvshome.org. They all build with the standard "sh ./configure", "make", and "make install". Note: at this point it isn't clear whether we will initailly send you a compressed, tarred file for the Coral source or have you extract it directly from our CVS repository ... 3. Other useful things: bash, git, ssh. Bash (a shell) and git (for Gnu Interactive Tools) are both available from gnu. ssh (Secure Shell) is available from www.ssh.com. Ssh is sort of a secure version of ftp and telnet rolled into one. Around here, our computer security folks are increasingly reeluctant to allow us to use either telnet or ftp because both of those tools allow login names and passwords across the network "in the clear" ... anyone with a sniffer can easily learn lots of login names and passwords that way. Assuming that we will likely ultimately need accounts on your machine to help configure/troubleshoot your system, you should probalby have ssh installed. Note: we actually run both the older ssh1 as well as the newer ssh2 protocols because there are places on campus here that only run ssh1. If you can, you are better off getting and installing only ssh2 (we have version 2.4.0) ... both because it is easier to install and configure and because it is more secure. Oh yes, one other thing: stripcmt ... this is a simple program that strips comments out of C and java programs that we use. It is availabe from: 4. Editing environment for java source files. We use either emacs (available from GNU) or xemacs (available from www.xemacs.org) as our primary text editor. In addtion, we run a package called JDE (for Java Development Environment) on top of that. If you expect to be modifying source files ... and either you will or we will at your site ... it probalby makes sense to have a JDE/emacs or JDE/xemacs installation. JDE is availabe at www.sunsite.dk/jde You will see that this also requires the installation of several other packages including: elib, eieio, semantic and speedbar. This is actually the trickiest thing to install ... but the JDE web site gives pretty good instructions for installing it with either emacs or xemacs. Let me know if you have any problems getting this stuff installed and configured on your machine. Hopefully, having these tools and utilities available will make the "real installation" of Coral much less painless. If you find these instuctions and comments confusing, please let me know ... I'm hoping to use this as the "version 0" set of "HowTo Coral". Thanks, John From mbell at snf.stanford.edu Tue May 29 15:03:34 2001 From: mbell at snf.stanford.edu (Mike Bell) Date: Tue, 29 May 2001 15:03:34 -0700 Subject: question on running remote applications on Coral/sunray References: Message-ID: <3B141CB5.4180AB10@snf.stanford.edu> Brian, The only thing we have changed recently is the resolution of the sunray displays. It was RESOLUTION=1280x1024 and now it's. RESOLUTION=1024x768 This was done to support the new flat panel sunrays. We could test setting it back on a single sunray to see if this is the issue. Let me know if and when you want to do this. Thanks, Mike Brian Joseph Greene wrote: > Hi Mike, > I used to be able to run remote graphical applications on Coral by > ssh'ing into my group computer (cvd.stanford.edu) and then starting the > application (xterm as a simple example). For some reason I cannot do this > anymore. > When I try to run xterm, I get the error message that I don't have my > DISPLAY properly set. > $DISPLAY is set to cvd:10.0 on cvd, but on the terminal I open in Coral, > $DISPLAY is set to :18.0 (I'm sure this changes each time I log into > coral). Setting the display on cvd to sunray:18.0 kills xterm > immediately after starting it, and snf:18.0 doesn't seem to work either. > Do you know what settings I should make to get this to work? > > Thanks in advance, > Brian > _____________________________________________________________ > Brian J. Greene Office: CIS-X 128X > Applied Physics Dept, MC 4090 Phone: (650) 723-4194 ext 5 > Stanford University Pager: (650) 354-9687 > Stanford, CA 94305-4090 Fax: (650) 723-4659 From bmurray at snf.stanford.edu Wed May 30 08:50:28 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Wed, 30 May 2001 08:50:28 -0700 (PDT) Subject: httpd on rosen Message-ID: Mike and John, This morning for the first time I noticed that an httpd is running on rosen. I don't believe we need that running or should have it running. It's a security risk. Am I dreaming or has something changed on rosen? Here's the ps: [~] $ ps -ef | grep http root 369 1 0 May 16 ? 0:00 ./httpd -f /etc/httpd/conf/httpd.conf www 21943 369 0 23:33:39 ? 0:00 ./httpd -f /etc/httpd/conf/httpd.conf www 21301 369 0 17:24:01 ? 0:00 ./httpd -f /etc/httpd/conf/httpd.conf www 21376 369 0 18:30:03 ? 0:00 ./httpd -f /etc/httpd/conf/httpd.conf Thanks, Bill From bmurray at snf.stanford.edu Wed May 30 08:56:05 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Wed, 30 May 2001 08:56:05 -0700 (PDT) Subject: httpd Message-ID: Mike and John, I assumed that we had been providing our iors from snf, but that's not the case. We have been providing them from labadmin which is rosen. So that's why the httpd is running there. I guess there are some benefits in that this provides a single point of failure. We may wish to provide access to the iors from snf in the future. As much as I look at rosen, I'm surprised I've never noticed the httpd running. Bill From mbell at snf.stanford.edu Thu May 31 09:43:25 2001 From: mbell at snf.stanford.edu (Mike Bell) Date: Thu, 31 May 2001 09:43:25 -0700 Subject: Coral questions Message-ID: <3B1674AD.8D0A57A1@snf.stanford.edu> Bill and John, Dick Crane had a couple of questions this morning concerning Coral. 1) Tylan2 failed to change to a red light even though a shutdown had been done. However, when Dick did a clear and then another shutdown the light then changed green to red. I suggested this might be known problem that occurs when the database is inconsistent with the actual state of the machine. Is that correct? 2) When does history update to show what has occurred on equipment. Nightly, real-time or ? 3) In testing Tylan2 the machine became no longer enabled in the member's name. Is there a way for staff to enable a piece of equipment in a members name? If so how do they set the charges appropriately if they know what they should be? Thanks, Mike From bmurray at snf.stanford.edu Thu May 31 10:49:10 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Thu, 31 May 2001 10:49:10 -0700 (PDT) Subject: Coral questions In-Reply-To: <3B1674AD.8D0A57A1@snf.stanford.edu> Message-ID: See answers below. On Thu, 31 May 2001, Mike Bell wrote: > Bill and John, > > Dick Crane had a couple of questions this morning concerning Coral. > > 1) Tylan2 failed to change to a red light even though a shutdown had > been done. However, when Dick did a clear and then another shutdown the > light then changed green to red. I suggested this might be known problem > that occurs when the database is inconsistent with the actual state of > the machine. Is that correct? > The tylan2 had actually been shutdown. The client does not change the light to red until the shutdown event has been received. At certain times (around 7:30am and midnight), there appear to be heavy network or processor loads and the event service slows down for a short period of time (15 to 20 minutes). During these slowdowns (external to Coral), events may take up to 5 minutes to post to all clients. Everything is working fine. However, the client doesn't yet know that the shutdown has occurred. > 2) When does history update to show what has occurred on equipment. > Nightly, real-time or ? > History is update in real-time. However, an equipment enable does not become history until the machine is actually enabled. If someone leave a machine enabled for days (which happens), that history will not show until the machine is disabled. > 3) In testing Tylan2 the machine became no longer enabled in the > member's name. Is there a way for staff to enable a piece of equipment > in a members name? If so how do they set the charges appropriately if > they know what they should be? > Yes. Just type the members name into the window and choose the appropriate account and project to charge for the time. > Thanks, > > Mike > From bmurray at snf.stanford.edu Thu May 31 10:53:45 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Thu, 31 May 2001 10:53:45 -0700 (PDT) Subject: Coral questions In-Reply-To: Message-ID: Sorry. In a hurry to finish coding. Second answer should read. History is updated in real-time. However, an equipment enable does not become history until the machine is actually *disabled*. If someone leaves a machine enabled for days (which happens), that history will not show until the machine is disabled. Bill From mbell at snf.stanford.edu Thu May 31 12:52:09 2001 From: mbell at snf.stanford.edu (Mike Bell) Date: Thu, 31 May 2001 12:52:09 -0700 Subject: Coral questions References: Message-ID: <3B16A0E9.881EF913@snf.stanford.edu> Bill, Well maybe, but Dick just tried it again and on the first shutdown on tylan2 nothing happens (except you do see the shutdown message.) On the second shutdown the light changes green to red immediately (you get a 2nd shutdown message.) This happened in an identical fashion in both instances. Coincidence - perhaps? Also as far as other point, I think the issue is, "If so how do they set the charges appropriately if they know what they should be?" That is, you can change the name for the enable, but you have a limited selection of accounts and projects and perhaps not the one you want for the partiular member. Anyway, Dick said he would grab you sometime and show you. Mike Bill Murray wrote: > See answers below. > > On Thu, 31 May 2001, Mike Bell wrote: > > > Bill and John, > > > > Dick Crane had a couple of questions this morning concerning Coral. > > > > 1) Tylan2 failed to change to a red light even though a shutdown had > > been done. However, when Dick did a clear and then another shutdown the > > light then changed green to red. I suggested this might be known problem > > that occurs when the database is inconsistent with the actual state of > > the machine. Is that correct? > > > > The tylan2 had actually been shutdown. The client does not change the > light to red until the shutdown event has been received. At certain > times (around 7:30am and midnight), there appear to be heavy network > or processor loads and the event service slows down for a short period > of time (15 to 20 minutes). During these slowdowns (external to Coral), > events may take up to 5 minutes to post to all clients. Everything is > working fine. However, the client doesn't yet know that the shutdown > has occurred. > > > 2) When does history update to show what has occurred on equipment. > > Nightly, real-time or ? > > > > History is update in real-time. However, an equipment enable does not > become history until the machine is actually enabled. If someone leave > a machine enabled for days (which happens), that history will not show > until the machine is disabled. > > > 3) In testing Tylan2 the machine became no longer enabled in the > > member's name. Is there a way for staff to enable a piece of equipment > > in a members name? If so how do they set the charges appropriately if > > they know what they should be? > > > > Yes. Just type the members name into the window and choose the appropriate > account and project to charge for the time. > > > Thanks, > > > > Mike > > From bmurray at snf.stanford.edu Thu May 31 13:35:09 2001 From: bmurray at snf.stanford.edu (Bill Murray) Date: Thu, 31 May 2001 13:35:09 -0700 (PDT) Subject: Coral questions In-Reply-To: <3B16A0E9.881EF913@snf.stanford.edu> Message-ID: Mike, I've restarted the servers to make sure the cache and the database are synchronized. As far as accounts go, the *limited* selection of accounts and projects are the only projects and accounts that the member can legally charge to. If the account isn't on the list, the administrative staff should be notified so the list can be updated. If this is not enforced, we have an accounting nightmare at month end. The university accounting system will reject the complete months charges if it finds a single expired or illegal account. Bill On Thu, 31 May 2001, Mike Bell wrote: > Bill, > > Well maybe, but Dick just tried it again and on the first shutdown on tylan2 > nothing happens (except you do see the shutdown message.) On the second > shutdown the light changes green to red immediately (you get a 2nd shutdown > message.) This happened in an identical fashion in both instances. Coincidence > - perhaps? > > Also as far as other point, I think the issue is, "If so how do they set the > charges appropriately if they know what they should be?" That is, you can > change the name for the enable, but you have a limited selection of accounts > and projects and perhaps not the one you want for the partiular member. > > Anyway, Dick said he would grab you sometime and show you. > > Mike > > Bill Murray wrote: > > > See answers below. > > > > On Thu, 31 May 2001, Mike Bell wrote: > > > > > Bill and John, > > > > > > Dick Crane had a couple of questions this morning concerning Coral. > > > > > > 1) Tylan2 failed to change to a red light even though a shutdown had > > > been done. However, when Dick did a clear and then another shutdown the > > > light then changed green to red. I suggested this might be known problem > > > that occurs when the database is inconsistent with the actual state of > > > the machine. Is that correct? > > > > > > > The tylan2 had actually been shutdown. The client does not change the > > light to red until the shutdown event has been received. At certain > > times (around 7:30am and midnight), there appear to be heavy network > > or processor loads and the event service slows down for a short period > > of time (15 to 20 minutes). During these slowdowns (external to Coral), > > events may take up to 5 minutes to post to all clients. Everything is > > working fine. However, the client doesn't yet know that the shutdown > > has occurred. > > > > > 2) When does history update to show what has occurred on equipment. > > > Nightly, real-time or ? > > > > > > > History is update in real-time. However, an equipment enable does not > > become history until the machine is actually enabled. If someone leave > > a machine enabled for days (which happens), that history will not show > > until the machine is disabled. > > > > > 3) In testing Tylan2 the machine became no longer enabled in the > > > member's name. Is there a way for staff to enable a piece of equipment > > > in a members name? If so how do they set the charges appropriately if > > > they know what they should be? > > > > > > > Yes. Just type the members name into the window and choose the appropriate > > account and project to charge for the time. > > > > > Thanks, > > > > > > Mike > > > >