Message boards :
Number crunching :
Upload OK; Update/Download not happening.
Message board moderation
Author | Message |
---|---|
Qui-Gon Send message Joined: 15 May 99 Posts: 2940 Credit: 19,199,902 RAC: 11 |
I've looked through a number of posts here in Number Crunching, and I have not seen this, but I presume it has been happening to everyone: The number of "results" listed as "Ready to send" has been increasing all morning--it is presently up to more than 212,000. I can upload results, but when I try to update, I get a "no schedulers responded" message and nothing will download. Would someone please update the news to let us know what is happening. |
David@home Send message Joined: 16 Jan 03 Posts: 755 Credit: 5,040,916 RAC: 28 |
I've looked through a number of posts here in Number Crunching, and I have not seen this, but I presume it has been happening to everyone: We are all having similiar issues due to the problems with the scheduler server. An update from my system did get acknowledged recently so things are happening. After an outage the scheduler gets flooded with requests so all we can do is wait until things settle back to normal levels of activity. |
Qui-Gon Send message Joined: 15 May 99 Posts: 2940 Credit: 19,199,902 RAC: 11 |
We are all having similiar issues due to the problems with the scheduler server. An update from my system did get acknowledged recently so things are happening. After an outage the scheduler gets flooded with requests so all we can do is wait until things settle back to normal levels of activity. I have seen connection delays after an outage many times in the past, but this time, I can connect to upload, but I can't download any work units even though there are more than 200K ready to send. This is different, and I would still like to see the news section updated. |
Metod, S56RKO Send message Joined: 27 Sep 02 Posts: 309 Credit: 113,221,277 RAC: 9 |
I have seen connection delays after an outage many times in the past, but this time, I can connect to upload, but I can't download any work units even though there are more than 200K ready to send. This is different, and I would still like to see the news section updated. As I understand it, there are three stages involved and all of them are fairly independant from each other:
Metod ... |
Qui-Gon Send message Joined: 15 May 99 Posts: 2940 Credit: 19,199,902 RAC: 11 |
As I understand it, there are three stages involved and all of them are fairly independant from each other: From the Server Status page, is this description: Scheduler: Determines what work is going to be sent to/received from requesting clients. If this is off, you cannot send any results or receive any workunits. You have cited something about connecting to an "uploader", but the Server Status page lists no such device. It seems that the Scheduler controls uploads and downloads, according to the official description found on the Server Status page. The Status Page also shows that the scheduler is running and there are plenty of "results" to send, so there is no explanation for the inability of all of us to download the 212K+ results that are there. That's why I am asking for an update to the news column on the SETI hope page (or the technical news page). |
cjsoftuk Send message Joined: 3 Sep 04 Posts: 248 Credit: 183,721 RAC: 0 |
Because of recent website outages I have added the website status to the BOINC Project Status at http://boincprojectstatus.ath.cx. S@H status at http://boincprojectstatus.ath.cx/SS_SETI@Home.html Please use that website to the full extent! |
MikeSW17 Send message Joined: 3 Apr 99 Posts: 1603 Credit: 2,700,523 RAC: 0 |
Because of recent website outages I have added the website status to the BOINC Project Status at http://boincprojectstatus.ath.cx. See my comment in your other thread.... PS What is ARPS? {From your WEB site} |
cjsoftuk Send message Joined: 3 Sep 04 Posts: 248 Credit: 183,721 RAC: 0 |
Because of recent website outages I have added the website status to the BOINC Project Status at http://boincprojectstatus.ath.cx. @MikeSW17 If you are trying to get personal and try to take the mickey out of every single thing I try to do, I shall have you blacklisted in my forum prefs. If you wish to learn more about this idea of mine called ARPS then perhaps you had better wait several weeks until I have fixed up the website. To e-mail me about this, use arps@cjsoftuk.dyndns.org, but wait a few days until the server for my e-mail is back up. |
Qui-Gon Send message Joined: 15 May 99 Posts: 2940 Credit: 19,199,902 RAC: 11 |
cjmweb uk, Are you a Berkeley developer or administrator? Where did you get the idea that there is an "uploader"? I have posted a link to the official SETI/BOINC Server Status page for my post, but when MikeSW17 pointed out a problem with your link, which has information not found on the Berkeley Status Page (and incorrect information at that), you respond as if he has attacked you. All he is trying to do, and all I am trying to do, is get accurate (hopefully official) information about what is happening now. If you have accurate/official information, then please share it--if not, then to use your own words in another thread, "shut up" (no offense intended). |
MikeSW17 Send message Joined: 3 Apr 99 Posts: 1603 Credit: 2,700,523 RAC: 0 |
Because of recent website outages I have added the website status to the BOINC Project Status at http://boincprojectstatus.ath.cx. Actually I wasn't "taking the mick", dont be so sensitive. I simply observed that your status site doesn't reflect the experience I am presently experiencing with Berkeley connectivity. That got wondering if there was a geographical element to the connectivity issues, that led me to your profile, thence to your web site. If something is ment to be secret - don't publish it on the web. On the other hand if you do, don't be surprised when someone with an interest in DC picks-up on something you have on your site. |
John McLeod VII Send message Joined: 15 Jul 99 Posts: 24806 Credit: 790,712 RAC: 0 |
I believe that the scheduler controls the request for work and the report of that work. There is another server that holds the files to download for a particular WU. And there is a server (that is possibly the same as the download server) that gets the files that are uploaded. The upload and download are a fairly straight forward file copy. The reason that upload works better than download is that a download has a lot more data than an upload. BOINC WIKI |
Alinator Send message Joined: 19 Apr 05 Posts: 4178 Credit: 4,647,982 RAC: 0 |
cjmweb uk, One thing to keep in mind is the documentation on the SETI web pages is way out of date and virtually useless except in broad general terms. The truth is yes it's possible to connect to the undocumented "Upload Server", regardless of whether the Scheduler is up or not, contrary to the info you referred to. Also the Project Devs and Adms have made reference to the ULDL server here on the boards (although I don't have a link handy and don't feel like searching for the post). Alinator |
Qui-Gon Send message Joined: 15 May 99 Posts: 2940 Credit: 19,199,902 RAC: 11 |
One thing to keep in mind is the documentation on the SETI web pages is way out of date and virtually useless except in broad general terms. I'm not saying there is no "Upload Server", just that I have not seen anything official about it, and if that is the explaination for why I can upload, but not download (based on the name "Upload Server" the problem should be the opposite), I'd like something more official than some guy with a website tossing off some unsubstantiated opinion. Thus, my request for a news update. |
Alinator Send message Joined: 19 Apr 05 Posts: 4178 Credit: 4,647,982 RAC: 0 |
One thing to keep in mind is the documentation on the SETI web pages is way out of date and virtually useless except in broad general terms. Well keep in mind the semantics of up/download depend on the which end of the connection you're looking from. In this case based on the Web site documentation they're defining it from the clients viewpoint. The observation I made about the Upload Server come from yesterday before SAH had the big meltdown, they had the Scheduler and Transitioners offline (as indicated on the Status Page), yet I had result data files get uploaded from a couple of machines during this interval. Alinator |
Qui-Gon Send message Joined: 15 May 99 Posts: 2940 Credit: 19,199,902 RAC: 11 |
OK, this is what I have been asking for: [url=http://setiweb.ssl.berkeley.edu/tech_news.php]Technical News<a> Thanks for the information. |
cjsoftuk Send message Joined: 3 Sep 04 Posts: 248 Credit: 183,721 RAC: 0 |
cjmweb uk, Well, actually, I think this is absurd! There is the file called file_upload_handler which does just that! MikeSW17 is trying to point out that the page is not entirely accurate, but it is very clearly almost correct. |
Tigher Send message Joined: 18 Mar 04 Posts: 1547 Credit: 760,577 RAC: 0 |
The ULDL server was mentioned in this thread. http://setiweb.ssl.berkeley.edu/forum_thread.php?id=15485 By a project dev. Quote: *In the case of the ULDL server, it passed all tests at lower volume loads.* Well I have 3 dry CPUs and 2 more about to dry up. Uploads happen sporadically but downloads are dead. Cant get cpdn units for them as on those PCs the cpdn client keeps erroring out...can't join other projects as they are not taking new members ...or there projects are not consistent with how I want my cpu cycles to contribute to the end product. Looks like my spare CPU cycles will go to waste despite efforts not to have that happen. So...the philosphy of "just crunch another project if seti falls over" is somewhat fallacious this last weekend I think. |
Divide Overflow Send message Joined: 3 Apr 99 Posts: 365 Credit: 131,684 RAC: 0 |
Uploads are going across just fine because they don't require scheduler interaction to initiate. The second step of getting those uploaded results reported does involve the scheduler, and they won't show up in the project as being returned until your client checks in with the scheduler successfully. Downloads are hard to come by right now because the scheduler is so overburdened. It takes a response from the scheduler to your client's request for more work to download WU's to your host. There could be hundreds of thousands of WU's ready to send, but if the scheduler is congested, your client won't be able to get the project to send it anything. |
Qui-Gon Send message Joined: 15 May 99 Posts: 2940 Credit: 19,199,902 RAC: 11 |
Well, actually, I think this is absurd! There is the file called file_upload_handler which does just that! MikeSW17 is trying to point out that the page is not entirely accurate, but it is very clearly almost correct. cjmweb uk, I'm not trying to pick a fight with you, but you assume too much: 1) You seem to think I knew there is an "Upload Scheduler", but I have not read through the thousands of posts here in Number Crunching (and a search would not be much good since this server has been called an uploader, an upload/download server and a few other names); 2) You seem to think that I am aware of a file named the file_upload_handler, but this is the first time I have heard of it; and even if I knew of it, it may be related to or used by the scheduler, which is described on the official Status page; 3) You say that MikeSW17 says your page is not entirely accurate, then you seem to agree, saying it is "very clearly almost correct", which in my book means the same thing. Thank you for your attempt to help, but the developers/administrators have already updated the Technical News page. |
Alinator Send message Joined: 19 Apr 05 Posts: 4178 Credit: 4,647,982 RAC: 0 |
Hmmm, looks like they just opened up the throttles again. Anybody taking odds on how long till the next "boiler" explosion? :-) Alinator |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.