Message boards :
Number crunching :
Panic Mode On (20) Server problems
Message board moderation
Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 . . . 15 · Next
Author | Message |
---|---|
Pappa Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 |
I could be a bit glib. Over the years, I remember when the Classic Servers were taken offline for weeks to sort Candidates. So what is 24 hours? Right, wrong or indifferent. I am also aware of some things hapening in that the Staff are trying to "balance" the problem to return to normal. Many have posted information as to potential reasons and their thoughts. Seti currently has a 100Megabit connection down the hill to the to the UCB Campus where the Gigabit link comes in. That link will support 12500000 Bytes per second transfer. You can do the math for MultiBeam take 12500000/367000 and look at the number (that is in seconds). That tells you how many downloads can happen at the same time. What makes it more complicated is that when we look at Astropulse. That file size is 8196000, so you can take now take 12500000/8196000 (that is also in seconds). The calculator tells you the rest. Now come up with a mix that will fit in the download bandwidth. That is what the Feeder is supposed to do. Currently there is an imbalance that is causing issues. So the Real Panic is how does the Seti Staff nail down the cause and correct it? Fast Ethernet, the Rules apply. When you get over 90% of the link capacity things start going bad in a hurry. If you want, please go to a hardware store and puchase an 8 foot lenght of pipe with an inside diameter of 1/2 inch. Stop at Toy store and purchase a couple of bags of marbles (you need to be able to overfill the 8 foot lenght of pipe). Now when you get home open the bags of marbles and start shoving them though the pipe (downloads). When the pipe is Full, start trying to shove marbles up the other end. Please take a moment to inform us of the results of the experiment. Yes I know of the flaw, Astropluse Marbles are Larger than 1/2 Inch. So the presumption that you mention is that Matt is testing how many downloads that can happen at the same time without interfering with Uploads which includes Scheduler Requests (get more work) or the Result that is Patiently waiting... I see nothing wrong with that. This does not mention that Seti Beta also shares the same Bandwith. Which is TOTALLY Disabled. It all "used to work," now what happened and why? So while typing this I see that Someone (Matt) has added new images and the splitters are doing both MB and AP Regards Please consider a Donation to the Seti Project. |
OzzFan Send message Joined: 9 Apr 02 Posts: 15691 Credit: 84,761,841 RAC: 28 |
Also when i try downloading work from my office the download always fail. Can you help me with this.. Most likely your downloads are being blocked by your IT Admins. Make sure you have permission to run BOINC on those computers, then ask your company's IT staff for assistance since they should know their network configuration better than we would. |
Geek@Play Send message Joined: 31 Jul 01 Posts: 2467 Credit: 86,146,931 RAC: 0 |
Wow.........Tuesday shutdown occured and the recovery period is on. I am still unable to upload or download. My 4 boxes have over 1000 work units ready to upload. I turned off network access. No use hammering the upload server when it's not even listening and turned off. Still don't understand the logic in having the upload server off and only downloads working. At some point all those downloads are going to turn into uploads aren't they? I don't know. When I return from vacation I will have to make a decision about continueing on this project or not. Right now at this time.....the only reason I am crunching is out of respect for my wingman. There may still be a time in the near future where all my work is done and I am unable to upload and report it. Monday my computers will absolutely be shut down. After that we shall see. Boinc....Boinc....Boinc....Boinc.... |
Uli Send message Joined: 6 Feb 00 Posts: 10923 Credit: 5,996,015 RAC: 1 |
Geek, first off, have a great vacation. Secondly, I hope you will stick around with maybe one comp or so. Your help on the boards has been very much appreciated. Plus you have your 8th year anni coming up. Now back to your regularly scheduled program. Panic mode. Pluto will always be a planet to me. Seti Ambassador Not to late to order an Anni Shirt |
Cosmic_Ocean Send message Joined: 23 Dec 00 Posts: 3027 Credit: 13,516,867 RAC: 13 |
While it is annoying and frustrating not being able to upload, as others have said, there is very likely a logical, and simple explanation for it being disabled. One thing I would hope would be done when they do decide to allow uploads again would be to turn off the download servers for 12-24 hours so that all the uploads can go through as quickly as possible. Of course this will cause a lot of people to start complaining because they can't download new work, but hey, the majority of us can't anyway because of too many pending uploads. :p Linux laptop: record uptime: 1511d 20h 19m (ended due to the power brick giving-up) |
Kleineschay2 Send message Joined: 14 Aug 99 Posts: 97 Credit: 23,218,261 RAC: 0 |
I have about 20 computers working for SETI and I a out of work units ALL THE TIME! |
BarryAZ Send message Joined: 1 Apr 01 Posts: 2580 Credit: 16,982,517 RAC: 0 |
Look, the deal is, with all that computing power, share the wealth -- join other projects to reduce the dependence on SETI. SETI has, so it seems, more user computing power than it can readily handle, other projects will welcome you, and by having other projects, you might not get as frustrated with SETI -- which at this point has an 'up time' (measured by upload availability) somewhere in the high 70% range. There are of course other definitions of 'up time'. Since you've suggested the possibility of leaving SETI behind, that suggests the possibility of you working with other BOINC projects. I mean I realize there are some folks who, regardless of the available computing power they can share (be it 1 or 100 CPU/GPU), simply have no interest in other projects. Wow.........Tuesday shutdown occured and the recovery period is on. I am still unable to upload or download. My 4 boxes have over 1000 work units ready to upload. I turned off network access. No use hammering the upload server when it's not even listening and turned off. Still don't understand the logic in having the upload server off and only downloads working. At some point all those downloads are going to turn into uploads aren't they? |
BarryAZ Send message Joined: 1 Apr 01 Posts: 2580 Credit: 16,982,517 RAC: 0 |
|
Kleineschay2 Send message Joined: 14 Aug 99 Posts: 97 Credit: 23,218,261 RAC: 0 |
...and? Can we please just get SETI working? Too much to ask? I don't want to slove any other problems, I just want to find LGM. |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
...and? SETI@Home is basically three people. They're doing their best, I'm sure. |
darengosse Send message Joined: 8 Mar 06 Posts: 9 Credit: 1,045,896 RAC: 0 |
hello. Since more than 24 hours impossible to transfer the completed work. I do not include/understand because I continue to receive work. 20 WU finished and in a state of transfer. The newspaper of BOINC n' do not stop growing bigger. Always the same message: Internet access OK - project servers may Be temporarily down Thank you very much to inform of this message the server, in order to solve this problem.... |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13727 Credit: 208,696,464 RAC: 304 |
I've got enough work for the next 2 and a bit days. If the problems extend past that, then i guess i won't have any work left to process. *shrug* It's not the end of the world. Grant Darwin NT |
# Bob Ahlers # Send message Joined: 30 Mar 01 Posts: 18 Credit: 10,209,954 RAC: 0 |
If simplex is the main issue on the connection, a possible (work around) solution could be to switch upload and download servers on-line and offline with a window of 24 hours or so. This way the pipe is used only for one way traffic and all crunchers can download on day "1" and upload it all again on the next. This way Seti buys them self’s time to work on servers, connections, etc without annoying anyone to much. I understand that without sufficient funding this problem can not be solved. But also understand that 150.000+ active users can not be left waiting. Maybe it would be wise to think about a workaround before shutting down the system, without notifying SO MANY crunchers that invest power, hardware and time to help the search or even just do it for credits. http://www.youtube.com/watch?v=W8_Kfjo3VjU :-) Can someone reboot that server 3 times please, THE WEBSITE IS DOWN hehe |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13727 Credit: 208,696,464 RAC: 304 |
If simplex is the main issue on the connection, a possible (work around) solution could be to switch upload and download servers on-line and offline with a window of 24 hours or so. Even with the issues that have been going on, it's been possible to upload & download within a 24hour period, so i don't thenk there's any need for that option at this stage. At the moment the problem isn't one of too much traffic. It's just that the upload server isn't online. Grant Darwin NT |
Vipin Palazhi Send message Joined: 29 Feb 08 Posts: 286 Credit: 167,386,578 RAC: 0 |
I have been actively crunching for little over a year now and have been through lot many ups and downtimes. I usually dont post any comments regarding them, however, this one being exceptionally long, has prompted me to think. We all know that seti lacks the necessary resources to provide the voluteers with a 100% project uptime. To this cause, we have been contributing in terms of cash as well as hardware. It is evident from Matt's posts that he has to juggle his time between maintaining the servers, kicking it at times and doing the coding for NTPCKR and other needed software (+ god knows what else). And we cannot forget that he and the other two are humans too and need rests and time with their families. I was wondering if it was possible for the volunteers to help seti by means other than just cash and hardware. That is, but helpting the staff in coding the necessary softwares, which will give Matt some more time to tinker with the server. The guys at Lunatics team are already helping with the optimized apps, but they could put that away for sometime and help Matt with his codes. What is the use of more optimised apps if the present ones are overloading the servers. I am no programmer, and neither do I know about the policies and procedures of Berkeley. This may not be the best solution, but every bit counts and I hope that some arrangement of this sort can be developed in order to ease things and make life easier for the seti boys. ______________ |
Kleineschay2 Send message Joined: 14 Aug 99 Posts: 97 Credit: 23,218,261 RAC: 0 |
Get them more coffee!! Can't hurt, might help. |
tullio Send message Joined: 9 Apr 04 Posts: 8797 Credit: 2,930,782 RAC: 1 |
I am running 6 BOINC projects. Most of them are having trouble with disk space and bandwidth. The reason is simple: too many users crunching too fast on their new CPUs, and the problem has worsened since CUDA has made its debut. But it is only at SETI that forums (fora at Oxford) have become almost hysterical with people ranting for their RAC, Something like this is also happening at AQUA, which also uses CUDA. Calm down a bit, Saint Peter will not take your RAC into consideration. Tullio |
# Bob Ahlers # Send message Joined: 30 Mar 01 Posts: 18 Credit: 10,209,954 RAC: 0 |
I agree with you; Vipin Would it help Seti if we (crunchers) provide rackspace and bandwidth in a datacenter nearby? This way only Seti needs to send data between the main office and the servers in that datacenter! Seti, what is your responce on this? |
WimTea Send message Joined: 15 Feb 02 Posts: 34 Credit: 909,865 RAC: 0 |
I feel some appreciation and support is needed for the Seti staff. I mean, a (b)leading edge, very complex environment with a shoestring budget for exploitation and STILL able to provide 99%+ availability (as in supplying with enough work) to 150,000+ users, is there a business or organization anywhere that can even remotely equal that?? Not to mention there is time found to further develop the environment! Kudos to you all ! (any IT project taking longer than a year tends to become a standing environment in my opinion) |
Speedy Send message Joined: 26 Jun 04 Posts: 1643 Credit: 12,921,799 RAC: 89 |
A post from Blurf regarding upload issues |
©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.