Message boards :
Number crunching :
Perhaps the Network Gurus can Explain...
Message board moderation
Author | Message |
---|---|
MikeSW17 Send message Joined: 3 Apr 99 Posts: 1603 Credit: 2,700,523 RAC: 0 |
In the recent data-storage hangs, when no data could be uploaded or downloaded, the traffic graphs for packets and bits/s both dropped to only around 50% of normal use. Why so much traffic if no actual data was being sent/received? Is this effectively showing that the communications protocol has a 50% overhead? (Or should that be 100% overhead???) |
Tigher Send message Joined: 18 Mar 04 Posts: 1547 Credit: 760,577 RAC: 0 |
In the recent data-storage hangs, when no data could be uploaded or downloaded, the traffic graphs for packets and bits/s both dropped to only around 50% of normal use. Hi Mike As I understand it there are classic users on the same routers and circuit. So we see them plus our failed attempts to make contact. Cheers Ian |
MikeSW17 Send message Joined: 3 Apr 99 Posts: 1603 Credit: 2,700,523 RAC: 0 |
In the recent data-storage hangs, when no data could be uploaded or downloaded, the traffic graphs for packets and bits/s both dropped to only around 50% of normal use. Ah! Good point - I had completely forgotten about Classic Thanks |
W-K 666 Send message Joined: 18 May 99 Posts: 19059 Credit: 40,757,560 RAC: 67 |
I noticed that as well but they did send out some information, don't know which connection it uses but my credits and RAC were updated in Boinc Manager for Seti during that period. Andy |
MikeSW17 Send message Joined: 3 Apr 99 Posts: 1603 Credit: 2,700,523 RAC: 0 |
Oh well, sorry for my dumb question ;( Yet, it does reveal a couple of interesting facts. Namely BOINC seems to be doing almost exactly as much work as Classic does. And, network capacity should not be a issue when (eventually) Classic goes off. |
Tigher Send message Joined: 18 Mar 04 Posts: 1547 Credit: 760,577 RAC: 0 |
Oh well, sorry for my dumb question ;( Hmmm well I guess the jury is out on that one. I have suggested to UCB that they model the recovery time when they have a few million users all trying to recover from a crash. The reason I say this is that I had, an hour or so ago, about 60 results to go back therefore I am asking for 60 new WUs. Most classic users had one WU and no cache....so when they went down all that was waiting was a request for one WU. With us.....well I am waiting for 60 right now. So I am 60 times the workload I was when I was a classic user. I see an exponetial decline in performance when it comes to recovery with 2 or 3 or 4 million users. I think the demands will be very much higher. I would hazard a guess that it will take weeks to recover - hope I'm wrong! We will have to wait and see I suppose but I would have thought some young budding mathematician would leap at this and give us a view. |
Celtic Wolf Send message Joined: 3 Apr 99 Posts: 3278 Credit: 595,676 RAC: 0 |
well I am waiting for 60 right now. So I am 60 times the workload I was when I was a classic user. I see an exponetial decline in performance when it comes to recovery with 2 or 3 or 4 million users. I think the demands will be very much higher. I would hazard a guess that it will take weeks to recover - hope I'm wrong! We will have to wait and see I suppose but I would have thought some young budding mathematician would leap at this and give us a view. 60!!! I sure hope you have more then one computer crunching.. If not I'd have to ask how far out is your connect time.. I'd rather speak my mind because it hurts too much to bite my tongue. American Spirit BBQ Proudly Serving those that courageously defend freedom. |
Tigher Send message Joined: 18 Mar 04 Posts: 1547 Credit: 760,577 RAC: 0 |
well I am waiting for 60 right now. So I am 60 times the workload I was when I was a classic user. I see an exponetial decline in performance when it comes to recovery with 2 or 3 or 4 million users. I think the demands will be very much higher. I would hazard a guess that it will take weeks to recover - hope I'm wrong! We will have to wait and see I suppose but I would have thought some young budding mathematician would leap at this and give us a view. Hmmmmm...well I got 3 HT P4s, one P4 and one P3. 10 days for all but the P3 - laptop 20 Gis disc - which is 5 days. Cable so it connects whenever. |
Pascal, K G Send message Joined: 3 Apr 99 Posts: 2343 Credit: 150,491 RAC: 0 |
well I am waiting for 60 right now. So I am 60 times the workload I was when I was a classic user. I see an exponential decline in performance when it comes to recovery with 2 or 3 or 4 million users. I think the demands will be very much higher. I would hazard a guess that it will take weeks to recover - hope I'm wrong! We will have to wait and see I suppose but I would have thought some young budding mathematician would leap at this and give us a view. Never has and I bet never will be millions of users on at the same time, that millions figure is for all the users that have processed at least one WU over the last 6 years, I think the highs number on was just under 6 hundred fifty thousand and again that was over a period of time and they had to crunch at least one WU......I will see if I can find that link.....AGAIN. Classic user stats Semper Eadem So long Paul, it has been a hell of a ride. Park your ego's, fire up the computers, Science YES, Credits No. |
John McLeod VII Send message Joined: 15 Jul 99 Posts: 24806 Credit: 790,712 RAC: 0 |
well I am waiting for 60 right now. So I am 60 times the workload I was when I was a classic user. I see an exponential decline in performance when it comes to recovery with 2 or 3 or 4 million users. I think the demands will be very much higher. I would hazard a guess that it will take weeks to recover - hope I'm wrong! We will have to wait and see I suppose but I would have thought some young budding mathematician would leap at this and give us a view. Not quite. That is for all of the users that ever signed up - processed a WU or not. About half never finished the first one. BOINC WIKI |
Pascal, K G Send message Joined: 3 Apr 99 Posts: 2343 Credit: 150,491 RAC: 0 |
well I am waiting for 60 right now. So I am 60 times the workload I was when I was a classic user. I see an exponential decline in performance when it comes to recovery with 2 or 3 or 4 million users. I think the demands will be very much higher. I would hazard a guess that it will take weeks to recover - hope I'm wrong! We will have to wait and see I suppose but I would have thought some young budding mathematician would leap at this and give us a view. Edited for content, Fuzzy Now if you are talking host that is a different, so just for the argument let's use and average of 20 WUs per host to upload after the sever is back online.ANd say that all 279,696 are still active so we hosts times WUs = 5,593,920 WUs to load. Now I suck at math but that looks like a good reason to stay off the update button and be patient........ XML date: 2005-06-08 04:54:01 GMT 1 days 16:54:11 old Totals: Last Day: Users 121,281 0 Hosts 279,696 0 Teams 16,785 0 Countries 199 0 Total Credit 1,696,563,712 0 Link below Classic Stats link Semper Eadem So long Paul, it has been a hell of a ride. Park your ego's, fire up the computers, Science YES, Credits No. |
©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.