Posts by David Emigh

41) Message boards : Number crunching : minirosetta v1.19 bug thread (Message 52982) Posted 11 May 2008 by David Emigh Post: I've got a "Compute Error" on 76,000+ seconds of CPU time. Reason: Access Violation (0xc0000005) at address 0x005C3051 write attempt to address 0x00000024 There is a large and detailed debugger report.
42) Message boards : Number crunching : Stuck Jobs and Access Violations ... Expected Behavior (Message 52964) Posted 10 May 2008 by David Emigh Post: We're on mini 1.19 now, and it is far superior, IMHO, to mini 1.15
43) Message boards : Number crunching : minirosetta v1.19 bug thread (Message 52900) Posted 7 May 2008 by David Emigh Post: Here is an access violation error after 68,000+ seconds of CPU time: Reason: Access Violation (0xc0000005) at address 0x005C3051 write attempt to address 0x00000024 There is a large and detailed debugger message.
44) Message boards : Number crunching : minirosetta v1.15 bug thread (Message 52821) Posted 1 May 2008 by David Emigh Post: I suspect the focus of development energy has already shifted to 1.16, but if there is any information still to be gleaned from the failures of 1.15, the data in this post may be significant. I changed the runtime preference for the two computers that I've been whining about for several days now, decreasing it from 24 hours to 3 hours. The result was dramatic. On one system, my success rate with mini 1.15 went from 0% to 100%. That computer is now 5/5 with the formerly un-runnable application. link to host The other system is still struggling, but the type of error reported has changed from "compute error" to "validate error." link to host
45) Message boards : Number crunching : minirosetta v1.15 bug thread (Message 52815) Posted 1 May 2008 by David Emigh Post: I am 0/5 now on mini 1.15 workunits with these two computers: hostid=623950 hostid=663412 Both have successfully completed mini 1.15 workunits on RALPH. {...} Make that 0/8 now. Neither of the two above linked computers, normally stable, reliable crunchers, has been able to successfully complete a mini 1.15 workunit since that application version was released on Rosetta. I have set the runtime preference for the two computers down to 3 hours, since that time frame seems to have worked for my wingmen on all of these failed tasks.
46) Message boards : Number crunching : minirosetta v1.15 bug thread (Message 52814) Posted 30 Apr 2008 by David Emigh Post: Finally, a success with mini 1.15! This computer, hostid=599043, made it to the end of a 24 hour run on resultid=159274026. It may be important to note that this is a 64 bit processor with 4GB RAM and 8GB of swap space...
47) Message boards : Number crunching : minirosetta v1.15 bug thread (Message 52794) Posted 30 Apr 2008 by David Emigh Post: I am 0/5 now on mini 1.15 workunits with these two computers: hostid=623950 hostid=663412 Both have successfully completed mini 1.15 workunits on RALPH. I am waiting to see how my wingman does with failure number five before taking any further action. My wingmen, on the four tasks that have reported successful after my failures, were: two Linux boxes, one WinXp box, and a Darwin box. One thing they all had in common was run times of less than 11,000 seconds, so are probably set for a 3 hour runtime preference. Have any other users had problems with mini 1.15 that seem related to long runtime preferences?
48) Message boards : Number crunching : minirosetta v1.15 bug thread (Message 52789) Posted 29 Apr 2008 by David Emigh Post: Two more errors to report, both include large and detailed debugger reports. resultid=159151381 resultid=159146976 An observation: In every case (these two, and the the preceding two, reported earlier in this thread) my "wingman" ran these tasks successfully. But in all four cases, my wingman's runtime preference was apparently much shorter than my own. Coincidence? I don't know. The sample size is small, but the correlation is perfect. I ran many successful mini 1.15 workunits on RALPH, where my runtime preference is set to the minimum. So far, I am 0/4 with mini 1.15 workunits on Rosie, where my runtime preference is set to the maximum. I would be curious to know if this relationship has been observed by any other crunchers.
49) Message boards : Number crunching : Problems with version 5.96 (Message 52777) Posted 28 Apr 2008 by David Emigh Post: Why does Rosetta keeps in memory, although I set it "off"??? Have you performed an update from the client since you last changed your preferences?
50) Message boards : Number crunching : minirosetta v1.15 bug thread (Message 52737) Posted 26 Apr 2008 by David Emigh Post: Two "Compute Errors" to report, both with large and detailed debugger messages. resultid=158425798 resultid=158582229 Two different computers had the errors linked above. Both computers had successfully run 1.15 tasks for RALPH@home.
51) Message boards : Number crunching : Huge splurge of client errors since reseting (Message 52643) Posted 21 Apr 2008 by David Emigh Post: {...} Can anyone confirm this, or point to me where I can seek confirmation? You might start by contacting Adam Bazinet, the originator of the post.
52) Message boards : Number crunching : looking for your ideas on a new desktop or new laptop for RAH and general use (Message 52400) Posted 12 Apr 2008 by David Emigh Post: I don't actually own one of these laptops, but I have done a lot of business with the company that sells them, and I have never been disappointed.
53) Message boards : Rosetta@home Science : DSN @ Home : Project cannot launch : no USB CD-ROM drive or MacMini to replace MiniPC (Message 52218) Posted 3 Apr 2008 by David Emigh Post: 4. ? Things I have not considered. Put the MiniPC up for auction on eBay. Set a reserve price equal to the purchase price as stated in the URL you provided to The Linux Store. If you do this, I will bid on the computer. My eBay ID is davide405. Please feel free to check my profile and feedback. I am not promising to be the purchaser of the computer, only that I will bid on it to meet (or exceed) the reserve price as specified above. If someone else outbids me, we all win. I have a homebuilt computer, tinyhome , that uses the same motherboard. It had Ethernet connection issues also, but I was able to resolve them, and am confident I can repeat the process.
54) Message boards : Number crunching : FAQ revamp - your comments welcome (Message 51941) Posted 14 Mar 2008 by David Emigh Post: I am interested. What might be entailed?
55) Message boards : Rosetta@home Science : T FLOPS (Message 51924) Posted 13 Mar 2008 by David Emigh Post: This might have something to do with it =)
56) Message boards : Number crunching : Not getting enough work... (Message 51811) Posted 5 Mar 2008 by David Emigh Post: {...}It goes through one every couple hours so it should have more like 100 on deck... {emphasis added} It appears that you have not adjusted your runtime preference from the default. The project gains equal scientific information from one WU that runs 8 hours as it does from four WUs that run 2 hours each, with the added (and significant) benefit of one-fourth of the internet traffic. The maximum allowed runtime is 24 hours, and all of my computers are set to that length. If you choose to adjust your runtime preference, do so gradually, adding a few hours a day until you reach your target. The BOINC manager needs time to "learn" how to manage your workload.
57) Message boards : Number crunching : Pentium II - Client 5.8.15 - 400Credits/Day (Message 51680) Posted 27 Feb 2008 by David Emigh Post: As far as I can tell, the longer the runtime the better, as far as Rosetta is concerned. However, be sure to make adjustments to your preferred runtime gradually. I suggest you add a few hours each day until you get to 24 hours.
58) Message boards : Number crunching : Best operating system for crunching (Message 51679) Posted 27 Feb 2008 by David Emigh Post: {...} Most importantly I would like an OS that is actually simple to use. I really wish a Linux Distro could just have the same type of "ease of installation" that Windows uses. You know what I am talking about; I can download any file, then double click said file, follow the on-screen instructions, and then I can start using the program. {...} There are at least three Linux distros that meet your requirements: Linspire, Ubuntu, and gOS. Using the synaptic package manager in gOS, I had BOINC installed, and Rosetta attached and running on my little linux box in less time than it has taken to make this post.
59) Message boards : Number crunching : Seriously, what for? (Message 51484) Posted 18 Feb 2008 by David Emigh Post: {...} But what exactly is my point you ask, right? {...} {emphasis added} I've read your post at least four times, and I'm still wondering.
60) Message boards : Number crunching : Ubuntu "boinc" user (Message 51453) Posted 17 Feb 2008 by David Emigh Post: I had a dual boot box once, this one, but eventually decided not to go that way. This is my experience with dual boot: Once you go dual boot, the G.R.U.B. owns your boot sector. You must always boot through the G.R.U.B. Supposedly, it is possible to change the default OS (i.e. WinXp instead of Ubuntu), but I never succeeded at that. As to installing BOINC on a linux box: Definitely use the Synaptic Package Manager. Trying to install BOINC the "old way" is not at all easy or intuitive for those not familiar with Linux. To escape dual boot land, I had to reformat the hard drive... fortunately, I had a spare activation on my WinXp license.

Previous 20 · Next 20