From komarek at cmu.edu Tue Jun 1 12:59:33 2004 From: komarek at cmu.edu (Paul Komarek) Date: Tue, 01 Jun 2004 12:59:33 -0400 Subject: [auton-users] quota issues on bigpapa Message-ID: <40BCB5F5.3040903@cmu.edu> Hello, We recently observed a quota malfunction on bigpapa for one user. If you find you cannot create new files or directories, let us know at admin at autonlab.org. For instance, gmake might fail with warnings from the assembler about being unable to create Linux...fast/foo.o. This problem is easily fixed in the short term (for NIPS). We'll address the longer term cause and remedy as soon as possible. Alternative communication methods: Paul: AIM: scaleusrubberdog IRC: #auton on irc.freenode.net Mobile: 412.983.1284 Jacob: AIM: jacobmj1 IRC: #auton on irc.freenode.net Mobile: 831.524.0666 If you can afford to wait a few hours, send email to admin at autonlab.org. -Paul From komarek at cmu.edu Tue Jun 1 14:33:53 2004 From: komarek at cmu.edu (Paul Komarek) Date: Tue, 01 Jun 2004 14:33:53 -0400 Subject: [auton-users] reservation request for Paul on lop6 Message-ID: <40BCCC11.9060105@cmu.edu> Hello, I would like to reserve lop6 for the rest of this week. Are there any objections? It is possible I will release it early, as I have only a few experiments to run. -Paul From komarek at cmu.edu Tue Jun 1 15:03:04 2004 From: komarek at cmu.edu (Paul Komarek) Date: Tue, 01 Jun 2004 15:03:04 -0400 Subject: [auton-users] reservation request for Paul on lop? In-Reply-To: <40BCCC11.9060105@cmu.edu> References: <40BCCC11.9060105@cmu.edu> Message-ID: <40BCD2E8.8020802@cmu.edu> Oops, lop6 is taken. Well, I'd like to reserve a lop, I don't care which. Any suggestions? All but lop6 and loq1 were busy when I checked. -Paul Paul Komarek wrote: > Hello, > > I would like to reserve lop6 for the rest of this week. Are there any > objections? It is possible I will release it early, as I have only a > few experiments to run. > > -Paul > From komarek at cmu.edu Tue Jun 1 15:06:08 2004 From: komarek at cmu.edu (Paul Komarek) Date: Tue, 01 Jun 2004 15:06:08 -0400 Subject: [auton-users] reservation request for Paul on lop? In-Reply-To: <40BCD2E8.8020802@cmu.edu> References: <40BCCC11.9060105@cmu.edu> <40BCD2E8.8020802@cmu.edu> Message-ID: <40BCD3A0.60802@cmu.edu> Oops again, Ajit nicely told me I can use lop6. If there are any objections, please let me know! -Paul Paul Komarek wrote: > Oops, lop6 is taken. Well, I'd like to reserve a lop, I don't care > which. Any suggestions? All but lop6 and loq1 were busy when I checked. > > -Paul > > Paul Komarek wrote: > >> Hello, >> >> I would like to reserve lop6 for the rest of this week. Are there any >> objections? It is possible I will release it early, as I have only a >> few experiments to run. >> >> -Paul >> > From komarek at cmu.edu Wed Jun 2 20:44:20 2004 From: komarek at cmu.edu (Paul Komarek) Date: Wed, 02 Jun 2004 20:44:20 -0400 Subject: [auton-users] reserving lop6 for Paul Komarek Message-ID: <40BE7464.7040709@cmu.edu> Hello, I'm going to take Ajit up on his offer from yesterday. I have transfered the reservation on lop6 to myself. -Paul Komarek From komarek at cmu.edu Thu Jun 3 14:01:18 2004 From: komarek at cmu.edu (Paul Komarek) Date: Thu, 03 Jun 2004 14:01:18 -0400 Subject: [auton-users] lop6 is free Message-ID: <40BF676E.9080700@cmu.edu> Ajit and I have released our reservations on lop6. -Paul Komarek From komarek at cmu.edu Thu Jun 3 16:06:08 2004 From: komarek at cmu.edu (Paul Komarek) Date: Thu, 03 Jun 2004 16:06:08 -0400 Subject: [auton-users] eeks! Need another reservation for NIPS! Message-ID: <40BF84B0.1050706@cmu.edu> Hi, Due to a recently discovered problem with my code, I need to rerun some tests. Can I reserve lop7 and loq1? -Paul Komarek From pgunn at cs.cmu.edu Fri Jun 4 14:06:08 2004 From: pgunn at cs.cmu.edu (Pat Gunn) Date: Fri, 4 Jun 2004 14:06:08 -0400 (EDT) Subject: [auton-users] Loom gone Message-ID: Hey all, Loom, the former general-purpose workstation, has been recruited for another task, and is no longer available for computation/logins. -- Pat Gunn Research/Systems Programmer, Auton Group, CMU From komarek at cmu.edu Sat Jun 5 11:55:06 2004 From: komarek at cmu.edu (Paul Komarek) Date: Sat, 05 Jun 2004 11:55:06 -0400 Subject: [auton-users] Beyond NIPS and Spring Madness Message-ID: <40C1ECDA.7020309@cmu.edu> Hello, It seems that NIPS and the other conferences that make up the Spring Madness are past. There are several jobs in the Auton computing lab that have been postponed due to the continual stream of conferences. We need to schedule a day of downtime, as well as make minor interruptions to service on the lops. Question: Is this week a bad week to interrupt lop service for a short time (2 hours)? Question: Is this week a bad time to schedule a day of downtime? Which days are bad? Congratulations to all the Auton Lab authors, you can probably afford to take (part of) the weekend off! ;-) -Paul From jmjoseph at andrew.cmu.edu Sat Jun 5 13:00:42 2004 From: jmjoseph at andrew.cmu.edu (Jacob Joseph) Date: Sat, 05 Jun 2004 13:00:42 -0400 Subject: [auton-users] Beyond NIPS and Spring Madness In-Reply-To: <40C1ECDA.7020309@cmu.edu> References: <40C1ECDA.7020309@cmu.edu> Message-ID: <40C1FC3A.4040909@andrew.cmu.edu> Wow. Paul beat me to it. I just wanted to indicate that any major downtime this week would not occur before Friday. If we see no one on a lop, and there are others available, we will probably take it offline for a short while without further notice. So, to be more specific, may we take down the fileserver beginning Friday afternoon, not to return possibly until late Saturday? -Jacob Paul Komarek wrote: > Hello, > > It seems that NIPS and the other conferences that make up the Spring > Madness are past. There are several jobs in the Auton computing lab > that have been postponed due to the continual stream of conferences. We > need to schedule a day of downtime, as well as make minor interruptions > to service on the lops. > > Question: Is this week a bad week to interrupt lop service for a short > time (2 hours)? > > Question: Is this week a bad time to schedule a day of downtime? Which > days are bad? > > Congratulations to all the Auton Lab authors, you can probably afford to > take (part of) the weekend off! ;-) > > -Paul From awd at cs.cmu.edu Mon Jun 7 12:08:07 2004 From: awd at cs.cmu.edu (Artur Dubrawski) Date: Mon, 7 Jun 2004 12:08:07 -0400 Subject: [auton-users] Artur's talk on Tuesday Message-ID: <000601c44ca9$a015be20$0200000a@artur> Dear Autonomous People, A few colleagues asked me to give them details of the talk time and place, and I thought I could share the info with everyone else as well. June 8th, 10am, NSH 3305 Artur ================================ Title: Marketing Autonomous Systems Abstract: How can we make systems that have the kind of useful autonomy acceptable to real world non technical users? This talk will begin with two examples of systems that try to achieve this: the first a successfully deployed autonomous hospital robot and the second a highly automated data analyzer. I will describe a project to build an automated delivery system called the Tug, which makes a good and not very common example of a success in developing an autonomous mobile robot to satisfy tangible needs of its users: the US hospitals. That was possible due to a diligent market analysis and a rapid adaptation to continually identified customer needs. I will discuss design and technical components that were involved in making an intrinsically complex product simple and intuitive to use by its target audience. Similar issues are often faced by scientists conducting applied research in other fields. In statistical data mining, we sometimes try to keep the users away from the intimidating complexity by equipping our solutions with a certain level of autonomy. I will discuss the theory and practice of Gaussian Mixture Trees, a combination of Dependency Trees and Gaussian Mixture Models, used to decompose complex, highly dimensional problems in a way that makes them generally more tractable and their solutions more intuitive and easier to comprehend by the end users. The talk will conclude with a higher level discuss of the issues of practical autonomy in systems used by non experts: this is a topic that I intend to research in the mid to long term. Bio: Artur Dubrawski received a Ph.D. in robotics and automation from the Institute of Fundamental Technological Research, Polish Academy of Sciences, and a M.Sc. in aircraft engineering from Warsaw University of Technology. Artur considers himself a scientist and a practitioner; he has been tainted with some real world entrepreneurial experiences. He had started up a small company which ended up being quite successful in integration and deployment of advanced computerized control systems and novel technological devices. He had also been affiliated with startups incorporated by others: Schenley Park Research, a data mining consultancy and a CMU off-spring, where he was a scientist; and (more recently) with Aethon, a company building robots to automate transportation in hospitals, where he served as a Chief Technical Officer. Artur returned to CMU last Summer to join the Robotics Institute's Auton Lab as a Project Scientist. He works on a range of applied data mining endeavors and teaches data mining to graduate students enrolled in the CMU MISM program (in fact, he has been doing that since 2000 as an adjunct). In his previous academic life, he worked mainly on machine learning approaches to mobile robot navigation and control, as well as on other applications of adaptive autonomous systems. In 1995/96 Artur spent a year at CMU as a visiting Fulbright scholar. He actively published scientific papers until his departure to the industrial world in 1999, and he certainly hopes to resume publishing any time now. -------------- next part -------------- An HTML attachment was scrubbed... URL: From dpelleg+ at cs.cmu.edu Tue Jun 8 09:49:30 2004 From: dpelleg+ at cs.cmu.edu (Dan Pelleg) Date: Tue, 8 Jun 2004 16:49:30 +0300 Subject: [auton-users] OS X build Message-ID: <16581.50154.260480.413853@localhost.localdomain> I'm trying to debug a build problem for h/kmeans for a user using Apple's OS X. I don't have such a machine, and to the best of my knowledge the lab does not have an OS X build/test machine. I'd appreciate it if you use OS X, and can spend 5 minutes looking at the report. If it's still a mystery I'll ask you to give me an account on your machine so I can test it further. --Dan From awd at cs.cmu.edu Wed Jun 9 13:25:49 2004 From: awd at cs.cmu.edu (Artur Dubrawski) Date: Wed, 09 Jun 2004 13:25:49 -0400 Subject: [auton-users] robobusiness workshop Message-ID: <1086801949.22939.14.camel@logo.auton.cs.cmu.edu> Pat suggested that I might want to let everyone else in the lab know about a series of workshops on how to turn a great (robotic or non-robotic) into a business: http://www.ri.cmu.edu/events/robobus.html#sem2 I attended the first one (in May), and it was a very interesting experience (btw, my former boss and founder of Aethon was among the speakers). The next meeting is on 17th of this month and the keynote speaker will be Don Jones, one of the most influential, resourceful and creative entrepreneurs headquartered in Pittsburgh, who happens to be a serious venture capitalist as well. If you ever dreamed about opening a business you should not miss this chance of learning useful things from seasoned practitioners. Artur From pgunn at cs.cmu.edu Wed Jun 9 14:22:57 2004 From: pgunn at cs.cmu.edu (Pat Gunn) Date: Wed, 9 Jun 2004 14:22:57 -0400 (EDT) Subject: [auton-users] lake being retired Message-ID: Hello all, The system "lake.auton.cs.cmu.edu" is being repurposed. I have backed up the Unix home directories (which lived in /usr0), and they should be available on BigPapa under data/casual-backups/lake-decommissioning by the time you get this. I have not backed up any data on the windows partitions -- if you want anything backed up on the windows side, please let me know ASAP -- when I hear back from Andrew and PaulHsiung, or by tomorrow (whichever happens later), the system will be cleaned. -- Pat Gunn Research/Systems Programmer, Auton Group, CMU From jmjoseph at andrew.cmu.edu Thu Jun 10 01:15:41 2004 From: jmjoseph at andrew.cmu.edu (Jacob Joseph) Date: Thu, 10 Jun 2004 01:15:41 -0400 Subject: [auton-users] Scheduled Downtime Fri-Sat Message-ID: <40C7EE7D.1050700@andrew.cmu.edu> Just a reminder: All cluster machines, including the file server, will be down for maintenance from Friday at 5:00pm to some time Saturday evening. As a reboot of all Opterons is planned for a bios upgrade, I cannot promise that any running jobs will survive. NFS will, of course be down. The website and email lists will remain up. As there were no objections to last Saturday's message about downtime, I'm assuming this will not be an inconvenience for anyone. Do let me know ASAP if plans have changed. -Jacob From jmjoseph at andrew.cmu.edu Fri Jun 11 14:59:32 2004 From: jmjoseph at andrew.cmu.edu (Jacob Joseph) Date: Fri, 11 Jun 2004 14:59:32 -0400 Subject: [auton-users] Final Downtime Countdown Message-ID: <40CA0114.3090304@andrew.cmu.edu> With 2 hours to go, I see that the lab load (http://www.autonlab.org/status) is again nearing 25. Of course, I know everyone planned their jobs accordingly, but just to be absolutely sure: I will disable access at 5pm and begin taking machines down shortly after. This includes all linux machines in the NSH A432 lab and NFS access for all. I can be reached by email, #AUTON, or AIM:jacobmj1 in event of emergency. -Jacob From jmjoseph at andrew.cmu.edu Sun Jun 13 03:47:57 2004 From: jmjoseph at andrew.cmu.edu (Jacob Joseph) Date: Sun, 13 Jun 2004 03:47:57 -0400 Subject: [auton-users] Lab is back Message-ID: <40CC06AD.1080507@andrew.cmu.edu> I'm happy to announce that the lab is now back in operation. The lops are again available, user files are back, and the hosted databases are again functional. A few things: -Given the nature of the file server changes, it's likely that you will have messages along the lines "Stale NFS Handle" or "Permission Denied". In the best case, you'll need to unmount and remount BigPapa. A sure-fire way of fixing this is to reboot your machine. Please do try this before reporting any major failures. -In the midst of all the work, lop4 began reporting memory related errors. I have taken it offline. -Despite the past few weeks of testing, significant changes to the firewall have been made. Please do report any troubles to admin at autonlab.org. -Jacob From dongryel at andrew.cmu.edu Tue Jun 15 15:03:10 2004 From: dongryel at andrew.cmu.edu (Dongryeol Lee) Date: Tue, 15 Jun 2004 15:03:10 -0400 (EDT) Subject: [auton-users] Reservation request for loq2 Message-ID: Hi everyone, I would like to reserve loq2 until 6 PM on Thursday. Let me know if there are any problems. Thanks, From jmjoseph at andrew.cmu.edu Wed Jun 16 17:02:27 2004 From: jmjoseph at andrew.cmu.edu (Jacob Joseph) Date: Wed, 16 Jun 2004 17:02:27 -0400 Subject: [auton-users] Reservation request for loq2 In-Reply-To: References: Message-ID: <40D0B563.4020404@andrew.cmu.edu> Loq2 has been reserved. Please let me know when you are in fact finished. -Jacob Dongryeol Lee wrote: > Hi everyone, > > I would like to reserve loq2 until 6 PM on Thursday. Let me know > if there are any problems. Thanks, From jmjoseph at andrew.cmu.edu Fri Jun 18 17:49:54 2004 From: jmjoseph at andrew.cmu.edu (Jacob Joseph) Date: Fri, 18 Jun 2004 17:49:54 -0400 Subject: [auton-users] Reservation request for loq2 In-Reply-To: References: <40D0B563.4020404@andrew.cmu.edu> Message-ID: <40D36382.1000601@andrew.cmu.edu> Dongryeol's reservation of loq2 has been extended through at least Monday. His jobs are taking longer than he expected. -Jacob From pgunn at cs.cmu.edu Thu Jun 17 10:10:39 2004 From: pgunn at cs.cmu.edu (Pat Gunn) Date: Thu, 17 Jun 2004 10:10:39 -0400 (EDT) Subject: [auton-users] Pat Gunn is leaving the Auton Lab Message-ID: Hey all, I'm leaving the Auton Lab, and, starting in July, will be working in the Act-R lab in the Psychology department, doing fMRI studies, programming, and doing sysadmin stuff, eventually to hopefully become a neuropsych grad student. I have enjoyed working with you, and if any of you want to grab lunch, I'm generally free. I will probably eventually lose my CS email address, so after about a month, you'll need to use my personal email, pgunn at dachte.org I will, of course, still be around on IM, and can be reached by (cell/only)phone at 412.480.3321 -- Pat Gunn Research/Systems Programmer, Auton Group, CMU From pgunn at cs.cmu.edu Fri Jun 18 10:43:46 2004 From: pgunn at cs.cmu.edu (Pat Gunn) Date: Fri, 18 Jun 2004 10:43:46 -0400 (EDT) Subject: [auton-users] Test Message-ID: Testing mailing list. -- Pat Gunn Research/Systems Programmer, Auton Group, CMU From jmjoseph at andrew.cmu.edu Fri Jun 18 23:22:12 2004 From: jmjoseph at andrew.cmu.edu (Jacob Joseph) Date: Fri, 18 Jun 2004 23:22:12 -0400 Subject: [auton-users] Lop4 Message-ID: <40D3B164.5050102@andrew.cmu.edu> I've brought lop4 back online, but with only 3GB of ram. I've found one of the modules to have gone bad and will probably receive the replacement later next week, requiring a few minutes of downtime at that point. In the meantime, I don't recommend running very time-sensitive benchmarks on this machine. Due to how the memory is segmented on the Opterons, each CPU has faster access to a single group of memory and slightly slower access to the other CPU's memory. Normally, we interleave accesses to normalize this difference(about 10%). However, with 2GB on one CPU and 1GB on the other, I'm not positive a slight difference in runtimes will not become apparent. I'll email again when I have a better estimate of the ram ETA. -Jacob From jmjoseph at andrew.cmu.edu Tue Jun 22 19:52:24 2004 From: jmjoseph at andrew.cmu.edu (Jacob Joseph) Date: Tue, 22 Jun 2004 19:52:24 -0400 Subject: [auton-users] Lop4 repaired Message-ID: <40D8C638.7040006@andrew.cmu.edu> Lop4 again has 4GB of memory. -Jacob From komarek at cmu.edu Tue Jun 29 17:11:58 2004 From: komarek at cmu.edu (Paul Komarek) Date: Tue, 29 Jun 2004 17:11:58 -0400 Subject: [auton-users] [Fwd: You should check before attempting to come to work tomorrow (Wed)] Message-ID: <40E1DB1E.7020306@cmu.edu> -------- Original Message -------- Subject: You should check before attempting to come to work tomorrow (Wed) Date: Tue, 29 Jun 2004 16:52:54 -0400 From: Andrew W Moore Reply-To: Andrew W. Moore To: auton-research at cs.cmu.edu CC: Andrew W. Moore This afternoon NSH was covered by plastic "Police Line Do Not Cross" banners with the perhaps over-caffeinated words BUILDING EMERGENCY emlazoned on the doors. You might not be allowed in even if you show up, so check your email first. I'll forward any news to auton-research at cs since I'm not sure whether everyone on this mailing list gets Jim Skees' announcements directly. Note that Wean Hall is fine, so if you have a wireless laptop you can work there! Andrew From dongryel at andrew.cmu.edu Tue Jun 29 19:27:18 2004 From: dongryel at andrew.cmu.edu (Dongryeol Lee) Date: Tue, 29 Jun 2004 19:27:18 -0400 (EDT) Subject: [auton-users] Reservation request for loq2 In-Reply-To: <40D36382.1000601@andrew.cmu.edu> References: <40D0B563.4020404@andrew.cmu.edu> <40D36382.1000601@andrew.cmu.edu> Message-ID: loq2 is free now. On Fri, 18 Jun 2004, Jacob Joseph wrote: > Dongryeol's reservation of loq2 has been extended through at least > Monday. His jobs are taking longer than he expected. > > -Jacob > > From sajid at cmu.edu Wed Jun 30 07:38:15 2004 From: sajid at cmu.edu (Sajid M. Siddiqi) Date: Wed, 30 Jun 2004 07:38:15 -0400 (EDT) Subject: [auton-users] power is back to normal In-Reply-To: <40E1DB1E.7020306@cmu.edu> References: <40E1DB1E.7020306@cmu.edu> Message-ID: <3624.141.158.67.21.1088595495.squirrel@webmail.andrew.cmu.edu> http://www.cmu.edu/cmnews/extra/040629_outage.html Sajid From jmjoseph at andrew.cmu.edu Wed Jun 30 16:54:56 2004 From: jmjoseph at andrew.cmu.edu (Jacob Joseph) Date: Wed, 30 Jun 2004 16:54:56 -0400 Subject: [auton-users] Discontinued machines Message-ID: <40E328A0.2020209@andrew.cmu.edu> I'm trying to investigate a few machines to see if we should be paying for CS support. Unless someone speaks up, I'd like to discontinue CS support for lego.auton.cs.cmu.edu. -Jacob