Message boards :
Number crunching :
Panic Mode On (80) Server Problems?
Message board moderation
Previous · 1 . . . 8 · 9 · 10 · 11 · 12 · 13 · 14 . . . 25 · Next
Author | Message |
---|---|
Cliff Harding Send message Joined: 18 Aug 99 Posts: 1432 Credit: 110,967,840 RAC: 67 |
- the root cause is a very poorly though through and implemented download scheduling algorithm that can do nothing apart from cause problems. I don't know what's broken, but I have stopped the scheduler from sending me any more AP tasks until this get cleared up. I have 10 on one machine, 7 that has been sitting here since yesterday and the scheduler won't send any more because of the tasks are stuck. There are 23 that are stuck on the other one that are holding up 99 other tasks from d/l'ing. Since everything here is being held up by the APs, I'm seriously considering aborting them just to get some work. The retry button on both machines are in severe pain because I am abusing them so much. I don't buy computers, I build them!! |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13727 Credit: 208,696,464 RAC: 304 |
The current round of sick download performance was happening before the APs started to move, so you can forget them as the culprit - the root cause is a very poorly though through and implemented download scheduling algorithm that can do nothing apart from cause problems. ? My hosts file is set to use the one server that behaves OK. Even that has been struggling since the network traffic maxed out. Before the network traffic maxed out downloads were coming down quickly. With the severe WU limits in place, and the reduced availability of AP work, it was possible that the network traffic would back off. For whatever reason, that hasn't happened. Grant Darwin NT |
rob smith Send message Joined: 7 Mar 03 Posts: 22188 Credit: 416,307,556 RAC: 380 |
You were the lucky one. Before the servers maxed out I was lucky to get any work, and when I did an MB was taking about 30 minutes to download, with numerous re-tries, now the servers are maxed out MBs are either taking 1 minute or, in the majority of cases 30 minutes with many re-tries. In "normal" times an MB takes about 10 seconds, and an AP a couple of minutes, with only the odd re-try - even when the servers are maxed those times only double. Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
James Sotherden Send message Joined: 16 May 99 Posts: 10436 Credit: 110,373,059 RAC: 54 |
Some weeks here at Seti the work units just fly from the downloader, AP included .Then we get weeks where they crawl at snail speed if your lucky. I see the GPU group has another drive for two new servers. Im just wondering if that will help. After all if your trying to push 10" diameter of water thru a 1" diameter pipe, how effective will that be? Is it hardware or software? or a combination of both. [/quote] Old James |
tbret Send message Joined: 28 May 99 Posts: 3380 Credit: 296,162,071 RAC: 40 |
My personal belief is that nobody knows. My fear is that nobody cares. My annoyance is that nobody seems to be trying to find-out (just for fun). My response is to "abandon all hope." |
shizaru Send message Joined: 14 Jun 04 Posts: 1130 Credit: 1,967,904 RAC: 0 |
Every now and then it sure does feel like it. Then eventually lucidity gets the better of me and I realize that (as far as my little cruncher is concerned) Seti enjoys 100% uptime. Well, at least for the past 2 years it has... Now obviously to make a statement like that I must be bending the rules a little bit. And by that I mean I'm not counting a couple weeks of server replacements or a multi day power outage or any other scenario the Lab has no control over (though I can't remember any others in the 2yr timeframe mentioned). Now I know this is no consolation to anyone with a PC quick enough to burn through their cache in a matter of hours. But I'd be lying if I said my little netbook hasn't been crunching pretty much non-stop for a fraction over 24 months now. All that means is, a lot of people must be doing something right. And it's not just the people in the lab that care. What about Richard who's always looking out for us? What about Brad who just posted a wall-of-text of accomplishments for the lab? That's just a couple of the people going above and beyond so things run smoothly for us. But not a definitive list of people who care. What about the Raistmers and the Jasons? The Joes, Claggys and Mikes? I'm sure that's just the tip of the iceberg. In fact I think the reality is that everybody here cares for the project. |
tbret Send message Joined: 28 May 99 Posts: 3380 Credit: 296,162,071 RAC: 40 |
So that my comment isn't misunderstood as being snarky: I did say my "fear" is that nobody cares, and by "nobody" I meant anyone who could do anything about server issues. I didn't even go so far as to say, my "observation" is that nobody cares. It is plain that a great number of people care a lot about this project. My slowest cruncher has only run out of work once in many months. Your point is well taken. |
Bernie Vine Send message Joined: 26 May 99 Posts: 9954 Credit: 103,452,613 RAC: 328 |
I did say my "fear" is that nobody cares, and by "nobody" I meant anyone who could do anything about server issues. I do not believe that Matt Jeff and Eric "don't care". Sorry that is going too far. I don't believe they have the time and resources to do what they want but to say the guys in the lab "don't care" is to me and insult. |
Speedy Send message Joined: 26 Jun 04 Posts: 1643 Credit: 12,921,799 RAC: 89 |
but to say the guys in the lab "don't care" is to me and insult. +1 |
David S Send message Joined: 4 Oct 99 Posts: 18352 Credit: 27,761,924 RAC: 12 |
I did say my "fear" is that nobody cares, and by "nobody" I meant anyone who could do anything about server issues. I think you're still not understanding him. He didn't say they don't care, he said his fear is that they don't care. There's a difference between a fear and a belief. David Sitting on my butt while others boldly go, Waiting for a message from a small furry creature from Alpha Centauri. |
Bernie Vine Send message Joined: 26 May 99 Posts: 9954 Credit: 103,452,613 RAC: 328 |
I did say my "fear" is that nobody cares, and by "nobody" I meant anyone who could do anything about server issues. I understand the words but why even hint that perhaps no one cares, because that is what that statement suggests. Anyone who has been around these boards should realise that "a lack of caring" is NOT the problem and it is disingenuous to even suggest it. Also coming from someone with a total of 95,733,064 and running 13 machines. It just strikes the wrong note with me. |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13727 Credit: 208,696,464 RAC: 304 |
I think you're still not understanding him. He didn't say they don't care, he said his fear is that they don't care. There's a difference between a fear and a belief. Belief or fear makes no difference- he is implying that they don't care. Of course if he was to read the donation thread he'd see that there are plans to address a couple of the major problems the project has. That they've gone to the effort to inform the fund raisers of their requirements would indicate to most people that they do care. Grant Darwin NT |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
I think I'm going to assume that the fact that you've got nothing better to do than argue semantic points from a post eight messages, two days and a maintenance outage ago is a good thing. It means that the project has (very quietly) started running so smoothly that you've all got nothing better to panic about :P |
QuietDad Send message Joined: 2 Oct 99 Posts: 83 Credit: 28,926,603 RAC: 59 |
What some forget, the original idea for distributed computing was that we donate our spare computer cycles for scientists to use for research purpuses. This was NEVER intended to be a "game" for points. |
W-K 666 Send message Joined: 18 May 99 Posts: 19048 Credit: 40,757,560 RAC: 67 |
I think I'm going to assume that the fact that you've got nothing better to do than argue semantic points from a post eight messages, two days and a maintenance outage ago is a good thing. It's at times like this when I think we should have a "Preparing to Panic" thread. |
Bernie Vine Send message Joined: 26 May 99 Posts: 9954 Credit: 103,452,613 RAC: 328 |
I think I'm going to assume that the fact that you've got nothing better to do than argue semantic points from a post eight messages, two days and a maintenance outage ago is a good thing. Quite correct I haven't been panicking for ages |
Rolf Send message Joined: 16 Jun 09 Posts: 114 Credit: 7,817,146 RAC: 0 |
Please, no panic! Panic is when people start loosing connection to reality - but we are living in reality. If "he" doesn't care (or "they"don't care), there are still two possibilities: Either he doesn't want to take care (his own decision) or he does not have the time to take care (somebody else decided). As an example take Matt's blog: http://setiathome.berkeley.edu/forum_thread.php?id=59157 Although it's nearly 3 years old, some parts could still be valid! I prefer the last Q-A-block: Q: What do you mean "you hit the limit?" Are you mad at us? |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 |
I've been receiving Short CPU tasks that insist they run Immediately. Plus, all my recently downloaded CUDA 23 tasks are...Shorties. I see an approaching Shortie Storm. Batten down the hatches, no need to panic, it's just a little Short thing. Unless... |
Bernie Vine Send message Joined: 26 May 99 Posts: 9954 Credit: 103,452,613 RAC: 328 |
|
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
Something's gone away.... Everything was still going fine about 5 hours ago when I went to bed. This morning I have a bunch of rigs that have dropped back to Einstein because they can't connect to the server to report and get new work. Whazzup? "Freedom is just Chaos, with better lighting." Alan Dean Foster |
©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.