Minirosetta 3.62-3.65
Message boards : Number crunching : Minirosetta 3.62-3.65
Previous · 1 · 2 · 3 · 4 · 5 . . . 8 · Next
| Author | Message |
|---|---|
 Chilean Send message Joined: 16 Oct 05 Posts: 711 Credit: 26,694,507 RAC: 0 |
I'm new to Rosetta@home in the last couple of days so the majority of the WUs that I've completed recently have ended up in a "Validation Failed" state. The bad part about it is the 5-6 hours of processing time for each job and we don't get any credit for the work. If I can provide additional error logs or information to fix the issue please let me know. I'd like to continue supporting this worthwhile project. You can change the running time in your preferences. I set mine to 2-3 hours for exactly this reason. 
|
|
Sid Celery Send message Joined: 11 Feb 08 Posts: 2122 Credit: 41,184,189 RAC: 10,001 
|
I'm new to Rosetta@home in the last couple of days so the majority of the WUs that I've completed recently have ended up in a "Validation Failed" state. The bad part about it is the 5-6 hours of processing time for each job and we don't get any credit for the work. If I can provide additional error logs or information to fix the issue please let me know. I'd like to continue supporting this worthwhile project. It's worth remembering that a clean-up job runs once a day that gives credit (equal to claimed credit) for validation errors, so you don't personally lose credit, even if it is delayed by a day. I also seem to recall it mentioned that the results submitted are still usable to the science bods. It just <looks> really bad - can't deny that.  
|
|
wyxchari Send message Joined: 27 Nov 14 Posts: 11 Credit: 85,318 RAC: 0 |
Rosetta should return to 3.61 to fix it. |
|
wyxchari Send message Joined: 27 Nov 14 Posts: 11 Credit: 85,318 RAC: 0 |
Since update 3.62, I can not run any task Rosetta. Works correctly: FiND@Home DENIS@Home SETI@Home Milkiway@home I using AMD Athlon 1700 with Windows XP SP3, 2Gb RAM. Download "minirosetta_3.62_windows_intelx86.exe". Is it okay if my CPU is AMD? Features: - MMX instructions - Extensions to MMX - 3DNow! technology - Extensions to 3DNow! - SSE / Streaming SIMD Extensions 
|
|
Ed Johnson Send message Joined: 9 Jun 06 Posts: 9 Credit: 4,738,577 RAC: 0 |
Task ID 756094939 failed under Linux, too. Hi Franziska, Any news on the FFD work units? |
|
Chilean Send message Joined: 16 Oct 05 Posts: 711 Credit: 26,694,507 RAC: 0 |
Since update 3.62, I can not run any task Rosetta. The "intelx86" part just refers to the architecture. With that said, you should look into investing into a cheap i3 computer in the future. It'd be WAY faster and use the same, if not less, energy. Don't get me wrong but you have one crappy computer. Not crappy, just old. I remember my first Athlon 2200+, they were so much better than the then P4. Any need for that? Well, he is running a very inefficient CPU by today's standards. Doesn't matter how slow, what matters it's how much power it draws for such slow computation. So he could be doing him a favor.
|
|
Irgendware Send message Joined: 3 May 15 Posts: 1 Credit: 871,903 RAC: 0 |
I am also using several PCs with AMD CPUs, but more modern ones - like AMD Turion 64 X2 or an AMD A10-7870K APU. But no Problems here, so far with minirosetta 3.62 |
|
Jim1348 Send message Joined: 19 Jan 06 Posts: 881 Credit: 52,257,545 RAC: 0 |
I just put this i7-4771 machine back on Rosetta since 3.62 came out, and no problems with 21 successes and no failures (Win7 64-bit). It looks like the worst is over. |
|
wyxchari Send message Joined: 27 Nov 14 Posts: 11 Credit: 85,318 RAC: 0 |
Has anyone tried Rosetta in AMD Athlon (without SSE2) with Windows XP? I have discarded software problems when installing cleaned Windows XP in VMware and the problem persists. I tend to think that Rosetta 3.62 fails with all Windows XP or Rosetta 3.62 needs SSE2 (like Google Chrome). The old computer is always on for network use (without specific personal use), so I take the opportunity to BOINC. For personal use, of course, I use one powerful.
|
Link Send message Joined: 4 May 07 Posts: 356 Credit: 382,349 RAC: 0 |
I tend to think that Rosetta 3.62 fails with all Windows XP or Rosetta 3.62 needs SSE2 (like Google Chrome). Roretta 3.62 runs here on WinXP without any issues. According to the application page it's supposed to be a generic i386 application, however exit code -1073741795 indeed indicates, that this is not the case. So they either need to build again a generic i386 app, or update the application info in the system, so the scheduler don't send it to computers, which can't execute it. . 
|
|
wyxchari Send message Joined: 27 Nov 14 Posts: 11 Credit: 85,318 RAC: 0 |
Thank you. Thank you. Thousand thanks. Finally, a coherent explanation. AMD Athlon has SSE, but not SSE2. That must be the problem. I switch to another project (WCG cancer) and I solved the problem. Thanks again. Greeting.
|
|
Sid Celery Send message Joined: 11 Feb 08 Posts: 2122 Credit: 41,184,189 RAC: 10,001
|
I tend to think that Rosetta 3.62 fails with all Windows XP or Rosetta 3.62 needs SSE2 (like Google Chrome). I was about to say it does run on such an AMD machine under XP but I noticed I'm using an AMD Athlon 64 X2 Dual Core Processor 3800+ on my backup machine which does have SSE2. That explains why things are fine here. Thanks. Hanging on by my fingertips!
|
Dr. Merkwürdigliebe Send message Joined: 5 Dec 10 Posts: 81 Credit: 2,657,273 RAC: 0 |
Just another example of why it's about time to finally discard support for ancient CPUs. They don't pack a punch. Everyone cares about some ancient Athlon, nobody (really) cares about the use of modern instruction sets like SSEx that have been around for ages now. For ages... AVX support not even mentioned. In a few years most people will have a CPU with such capabilities. But hey, why care about things like doubling the performance or so? Let's rather see how we can cater to people with ancient CPUs, next to nothing RAM and an OS that has already been phased out for good reasons. Sorry, admins, developers, this is the completely wrong way. Scientists in the HPC sector would laugh about this. |
|
Jim1348 Send message Joined: 19 Jan 06 Posts: 881 Credit: 52,257,545 RAC: 0 |
Also, the present application is 64-bits. Maybe the old CPUs are 32 bits? |
|
Chilean Send message Joined: 16 Oct 05 Posts: 711 Credit: 26,694,507 RAC: 0 |
I tend to think that Rosetta 3.62 fails with all Windows XP or Rosetta 3.62 needs SSE2 (like Google Chrome). Good god, I just hope the devs don't UNOPTIMIZE the app so it can run properly on 15+ year old CPUs. Rather, if they really want to support old CPUs (which is good for advertising), they should update the app info so BOINC doesn't send them to these CPUs, just like you stated.
|
|
[VENETO] boboviz Send message Joined: 1 Dec 05 Posts: 1994 Credit: 9,581,913 RAC: 7,911
|
Also, the present application is 64-bits. Maybe the old CPUs are 32 bits? App is 32 bit native. 64 is a wrapper |
|
[VENETO] boboviz Send message Joined: 1 Dec 05 Posts: 1994 Credit: 9,581,913 RAC: 7,911
|
Just another example of why it's about time to finally discard support for ancient CPUs. +1. But now admins are working on it |
|
sinspin Send message Joined: 30 Jan 06 Posts: 29 Credit: 6,574,585 RAC: 0 |
I am showing so far that many 3.62 tasks are encountering 'Validate Errors', though not all of them. Same here. 3 times 48 Hours wasted. Task 757088886 Task 755915384 Task 755915319 I will stop all FFD_* tasks until there is a true solution. Thanks. |
|
Sid Celery Send message Joined: 11 Feb 08 Posts: 2122 Credit: 41,184,189 RAC: 10,001
|
I am showing so far that many 3.62 tasks are encountering 'Validate Errors', though not all of them. These jobs have given credit, but limited to a flat 300 when the claimed amount is much more. Is this related to the run times being for 2 days? Someone highlkighted an issue with this runtime a few weeks ago, but I can't recall what the circumstances were. Point is, there seem to be a few consistent issues at play being reported in the last few days. Something is going wrong that needs an urgent look before more big crunchers get alienated.
|
|
sinspin Send message Joined: 30 Jan 06 Posts: 29 Credit: 6,574,585 RAC: 0 |
I have set the runtime to two days. That have two benefits. First of all, less of network traffic for download of new WUs. And, less of wasted time for finishing one WU and start of another. Means, more time for number crunching. In addition, it reduces the space on my disc which is needed for the project. |
Message boards :
Number crunching :
Minirosetta 3.62-3.65
©2024 University of Washington
https://www.bakerlab.org