Message boards : Number crunching : Problems and Technical Issues with Rosetta@home
Previous · 1 . . . 256 · 257 · 258 · 259 · 260 · 261 · 262 . . . 300 · Next
Author | Message |
---|---|
.clair. Send message Joined: 2 Jan 07 Posts: 274 Credit: 26,399,595 RAC: 0 |
Is anyone else getting fake downloads? Yes , I had lots of them when rosetta had work to do and the download server was `not running` and when I did get some to download ok they would not upload when crunched and timed out as `no response` And now the download server is `running` there is no work to download . funny old world . . . Two threads in number crunching forum :- Problen downloading work tasks. and Can't upload finished work units |
2am Send message Joined: 22 Jan 06 Posts: 1 Credit: 1,403,202 RAC: 0 |
Hello, for almost a month now, the downloads status is "Download: Retry in hh:mm:ss (project backoff: hh:mm:ss)" Nothing changed from a month ago, reinstalled and the issue stays the same. Platform is Windows 10. BOINC Manager Version: 7.16.11 (x64) wxWidget Version: 3.0.1 Any advice? Thanks |
.clair. Send message Joined: 2 Jan 07 Posts: 274 Credit: 26,399,595 RAC: 0 |
Hello, for almost a month now, the downloads status is "Download: Retry in hh:mm:ss (project backoff: hh:mm:ss)" Same as thousands of other Rosetta crunchers = were are stuffed until the admin fix the problem . Like - "Users in last 24 hours = 35" That's the total that are getting any work at the moment , otherwise known as something big is broken and it ain`t fixed , yet . |
Mr P Hucker Send message Joined: 12 Aug 06 Posts: 1600 Credit: 11,716,822 RAC: 13,035 |
Looks ok to me, they have very little work, it's been taken, and it's being returned. From the home page: Total queued jobs: 0 In progress: 4,026 Successes last 24h: 356 And the odd ones I've been getting download, process, then upload just fine. This is on 8 different Windows 11 machines ranging from 4 core to 24 core. AMD and Intel. |
.clair. Send message Joined: 2 Jan 07 Posts: 274 Credit: 26,399,595 RAC: 0 |
Yes , very little work to do , Just had a look through `top hosts` to see who gets work or not and it seem to be completely random . One of mine cant even reattach to the project. and another is unable to download the database and applications after reattaching . Both intel win7 . |
Jean-David Beyer Send message Joined: 2 Nov 05 Posts: 187 Credit: 6,365,440 RAC: 5,734 |
Yes , very little work to do , I looked at both my machines. Neither has any work. My main computer is running Red Hat Enterprise Linux release 8.7 (Ootpa). using an Intel(R) Xeon(R) W-2245 CPU @ 3.90GHz [Family 6 Model 85 Stepping 7] with 64 GBytes of RAM. On this one, I did a reset project because I could not see anything wrong with it, but it was full of junk. I did that only because it would not download anything even when the web site said there was work. It then tried to download the databae and applications (4 very large files), but has been unable to do it. It tries from time-to-time, but cannot do it. It is having no trouble with ClimatePrediction, World Community Grid, Einstein, MilkyWay, and Universe. It complains like this: Wed 18 Jan 2023 08:03:55 PM EST | Rosetta@home | Started download of rosetta_graphics_4.20_x86_64-pc-linux-gnu Wed 18 Jan 2023 08:03:55 PM EST | Rosetta@home | Started download of database_357d5d93529_n_methyl.zip Wed 18 Jan 2023 08:03:57 PM EST | | Project communication failed: attempting access to reference site Wed 18 Jan 2023 08:03:57 PM EST | Rosetta@home | Temporarily failed download of rosetta_graphics_4.20_x86_64-pc-linux-gnu: transient HTTP error Wed 18 Jan 2023 08:03:57 PM EST | Rosetta@home | Backing off 04:19:26 on download of rosetta_graphics_4.20_x86_64-pc-linux-gnu Wed 18 Jan 2023 08:03:57 PM EST | Rosetta@home | Temporarily failed download of database_357d5d93529_n_methyl.zip: transient HTTP error Wed 18 Jan 2023 08:03:57 PM EST | Rosetta@home | Backing off 04:22:30 on download of database_357d5d93529_n_methyl.zip Wed 18 Jan 2023 08:03:59 PM EST | | Internet access OK - project servers may be temporarily down. The other computer is a pipsqueak running Windows 10. It does not get any work from Rosetta either. GenuineIntel 11th Gen Intel(R) Core(TM) i7-1165G7 @ 2.80GHz [Family 6 Model 140 Stepping 1] Microsoft Windows 10 Core x64 Edition, (10.00.19045.00) |
rochester new york Send message Joined: 2 Jul 06 Posts: 2842 Credit: 2,020,043 RAC: 0 |
I had to discontinue Rosetta@home because i could no longer get any work units on my computers have been coming back every 3 or 4 months but have the same problem |
Mr P Hucker Send message Joined: 12 Aug 06 Posts: 1600 Credit: 11,716,822 RAC: 13,035 |
My main computer is running Red Hat Enterprise Linux release 8.7 (Ootpa). using anI have three running right now on my main Windows 11 24 core Ryzen, it got them 2.5 hours ago: https://boinc.bakerlab.org/rosetta/results.php?hostid=6181959. Resetting project on my other (almost identical) Windows 11 24 core Ryzen to see if it can download the main files.... https://boinc.bakerlab.org/rosetta/show_host_detail.php?hostid=6219929 1422 Rosetta@home Thu, 19/1/2023 10:33:58 AM Resetting project 1423 Rosetta@home Thu, 19/1/2023 10:35:26 AM update requested by user 1424 Rosetta@home Thu, 19/1/2023 10:35:33 AM Master file download succeeded 1425 Rosetta@home Thu, 19/1/2023 10:35:38 AM Sending scheduler request: Requested by user. 1426 Rosetta@home Thu, 19/1/2023 10:35:38 AM Not requesting tasks: don't need (CPU: not highest priority project; AMD/ATI GPU: ) 1427 Rosetta@home Thu, 19/1/2023 10:35:58 AM Scheduler request completed 1428 Rosetta@home Thu, 19/1/2023 10:35:58 AM Project requested delay of 31 seconds 1436 Rosetta@home Thu, 19/1/2023 10:36:46 AM update requested by user 1437 Rosetta@home Thu, 19/1/2023 10:36:50 AM Sending scheduler request: Requested by user. 1438 Rosetta@home Thu, 19/1/2023 10:36:50 AM Requesting new tasks for CPU 1439 Rosetta@home Thu, 19/1/2023 10:36:52 AM Scheduler request completed: got 0 new tasks 1440 Rosetta@home Thu, 19/1/2023 10:36:52 AM No tasks sent 1441 Rosetta@home Thu, 19/1/2023 10:36:52 AM Project requested delay of 31 seconds It did not attempt to download the main file as I'd hoped, to see if I could recreate your problem. Maybe it only does that if there are tasks available. Until such time, I won't know. The tasks I'm running may be using the data files downloaded yonks ago. Not sure how often they update them. I have an idea. Get hold of those two files from another machine. Here's the methyl database: (at half a GB, it will take a while to upload, see LHC below, I'll put the link here shortly). But as I only run Windows (on genuine licenses ROFL) I don't have the linux graphics file, only the windows equivalent. Somebody else in here may be able to provide it. I also use Squid on the main computer - this means all my 8 computers download the file once between them, then it's cached. I installed it to reduce the massive load on the LHC servers. I'm maxing out my 7Mbit uplink 24/7, and was getting a similar amount downloading (and the two were getting in the way of each other). At least now my LHC downloads are virtually nil. LHC programs are very stupid, every individual virtualbox task accesses LHC internally and downloads common files, so every single core on a 24 core computer is downloading the same thing 24 times for every task, then another 24 times when the next batch starts! |
Mr P Hucker Send message Joined: 12 Aug 06 Posts: 1600 Credit: 11,716,822 RAC: 13,035 |
Stupid useless forum doesn't let me edit an old post. Here's the file: https://www.dropbox.com/s/87mymss4dpmh9ru/database_357d5d93529_n_methyl.zip?dl=0 |
Jean-David Beyer Send message Joined: 2 Nov 05 Posts: 187 Credit: 6,365,440 RAC: 5,734 |
I have an idea. Get hold of those two files from another machine. Here's the methyl database: (at half a GB, it will take a while to upload, see LHC below, I'll put the link here shortly). But as I only run Windows (on genuine licenses ROFL) I don't have the linux graphics file, only the windows equivalent. Somebody else in here may be able to provide it. I do not want to get the four missing files independently. I want to be able to get them directly from the Rosetta web site. Because the fact that they cannot be downloaded that way is only a symptom of the real problem: the download server, even though now officially up, seems uable to be unable to download these things. |
Mr P Hucker Send message Joined: 12 Aug 06 Posts: 1600 Credit: 11,716,822 RAC: 13,035 |
I do not want to get the four missing files independently. I want to be able to get them directly from the Rosetta web site. Because the fact that they cannot be downloaded that way is only a symptom of the real problem: the download server, even though now officially up, seems uable to be unable to download these things.What an absurd attitude. I gave you a way to fix it and you can't be bothered. The simple fact remains I'm doing rosetta and you're not. You need to learn not everything in life is simple. |
Jean-David Beyer Send message Joined: 2 Nov 05 Posts: 187 Credit: 6,365,440 RAC: 5,734 |
I do not want to get the four missing files independently. I want to be able to get them directly from the Rosetta web site. Because the fact that they cannot be downloaded that way is only a symptom of the real problem: the download server, even though now officially up, seems uable to be unable to download these things. I do not think it absurd at all. I must still find out why it does not download those four files and fix that.. And if I get them from somewhere else, it will not get me any tasks, since according to the web site, there are none. And the reason I do not have them is that I reset the project. And the reason for that is it would not download tasks the same way it does not download those files. |
Mr P Hucker Send message Joined: 12 Aug 06 Posts: 1600 Credit: 11,716,822 RAC: 13,035 |
I do not think it absurd at all. I must still find out why it does not download those four files and fix that..If the problem is on their end, you can't fix anything, you just have to wait, and you might aswell get the files elsewhere and get on with crunching. If the fault is at your end, it's easier just to work round the problem. I can't see why the fault could be with your machine though, especially as others have said they have the problem too. Resetting the project should have cleared up any Boinc project problem, and since you haven't said you're experiencing any problems with other projects or your internet connection in general, I'd say it's the project that's broken. And if I get them from somewhere else, it will not get me any tasks, since according to the web site, there are none.They appear occasionally, either trickles from Robetta, or a day's worth from Steven Rettie. And the reason I do not have them is that I reset the project. And the reason for that is it would not download tasks the same way it does not download those files.Well that part is weird. What other projects do you run and are they all behaving ok? One thing you could try is find the address to download that file and see if your browser can download it. If it can't, it might give a useful error. |
ace28 Send message Joined: 16 Feb 10 Posts: 1 Credit: 526,688 RAC: 205 |
Thank you very much. Had the same problem, came to the forums, found the solution. Today is a good day =D |
Jean-David Beyer Send message Joined: 2 Nov 05 Posts: 187 Credit: 6,365,440 RAC: 5,734 |
I do not think it absurd at all. I must still find out why it does not download those four files and fix that.. I am having no problems getting tasks from ClimatePrediction, Einstein, World Community Grid, MilkyWay, and Universe. My Internet connection has had no problems in years. And if I get them from somewhere else, it will not get me any tasks, since according to the web site, there are none. I am having no problems getting tasks from ClimatePrediction, Einstein, World Community Grid, MilkyWay, and Universe. One thing you could try is find the address to download that file and see if your browser can download it. If it can't, it might give a useful error. I restored those four files from a backup. |
BubbleBoy Send message Joined: 5 May 14 Posts: 2 Credit: 634,699,331 RAC: 0 |
As I said months ago, the project has become trash. No official cares about it. It's abbandoned. You're wasting your time and money. |
Bryn Mawr Send message Joined: 26 Dec 18 Posts: 389 Credit: 12,070,320 RAC: 10,319 |
As I said months ago, the project has become trash. No official cares about it. It's abbandoned. You're wasting your time and money. In what way am I wasting my time and money? My preferred projects are Rosetta, WCG and CPDN with a couple of backups. They’re all set up to grab tasks from each project as and when they can - if Rosetta doesn’t give me work the others will and my machines are always busy. No waste of time and full return for money spent. |
harukaff Send message Joined: 15 Aug 07 Posts: 2 Credit: 433,932 RAC: 0 |
Rosetta's download server is having a very classic HTTPS configuration error: they forgot to put the intermediate certificate on the web server. Some HTTPS implementations won't actively try to fetch the intermediate certificate (the link is actually written in the certificate info), so they can't verify the certificate, leading to connection errors. If you enable the file_xfer_debuglog flag, you will see the error code is -184, which means cURL error. There is a very dirty workaround though, but don't do it on production systems - generally it should be safe but there is always a possibility this will come back and bite you: You can manually install the intermediate certificate as a trusted root certificate. After that, the system will be able to verify the certificate. If you can understand what the workaround is doing, I guess you already know how to do that. Otherwise, please know that this is inherently insecure, so I won't post a tutorial now. To anyone that can connect to the project team: please tell them about this issue and let them add the intermediate certificate to the web server. Shouldn't take more than a few minutes to fix. EDIT: fixed wonky English. |
Mr P Hucker Send message Joined: 12 Aug 06 Posts: 1600 Credit: 11,716,822 RAC: 13,035 |
HTTPS is a pain in the arse invented by worry warts. The internet worked a lot more reliably before it was invented. The end user should always have the option to ignore warnings about petty certificate dates. I mean seriously? 1 second out of date and suddenly it becomes dangerous? How? The site is somehow trusted for 10 years but not the day after that? What changed? |
harukaff Send message Joined: 15 Aug 07 Posts: 2 Credit: 433,932 RAC: 0 |
Encryption is almost a must though nowadays. You don't really want anyone on the network path to be able to just casually collect all the info you send through your network port. Not to mention that you actually download executables from BOINC projects that just directly runs on your machines. That said, the current PKI system is inherently flawed. There are ways to "fix" that but it's even more difficult to configure everything right. Guess it's just too hard to build verifiable trust in today's world. |
Message boards :
Number crunching :
Problems and Technical Issues with Rosetta@home
©2024 University of Washington
https://www.bakerlab.org