8 hours elapsed and I'm still not getting new work.

Is it saying "no work from project"??? That should be fixed now...

HM, it seems to flow with some problems; my boxes need 2 or 3 touches within a short time (3 - 5 Minutes), then they get work.

Same here, sometimes it takes 3-5 connects on it's own at 3-4 minute intervals but then I'll get the Requested Work.

I apologize for covering old ground, but my frustration level is rising and need to, um, vent just a little:

I use a queue size of 0.01 days (for reasons I don't care to explain) which has served me well for many weeks. But I continue to get "No Work From Project". I rationalized that if I changed queue size temporarily to 0.5 days and was lucky enough to actually get some work then I would like to "stock up". My next work request, however, gave me the infamous "Schedule Request Failed...with return value of 500", "No Schedulers Responded."

Great! Now I have 14 work units supposedly sent to me that I don't have onboard. I know Rosetta doesn't re-sync, but this would certainly help me.

If I'm destined to get so-called "ghost WUs" and to get "No Work From Project" msgs (a problem I don't experience with EaH, so far) then I don't see the profit in staying attached to Rosetta despite my strong wish to stay. And no, I'm not going to tweak MTU parameters, whatever they are. In other words, if only Rosetta is giving me a problem, then excise it.

The timing for these problems isn't great either.

Many folk are moving their boxen off FAD this week and looking for a new home project. They aren't going to be happy if they get the 'No Work Frm project' message as soon as they fire up Rosetta.

Let's hope the admins get this fixed soon.

---

The MTU "fix" seems to apply only to those who _consistently_ get "500"s, so I wouldn't recommend it for you anyway, and definitely it's a temporary patch regardless. The "out of work" bit is the only thing Rosetta can deal with, and I know they're working on it. Hopefully once they find and fix that problem, there will be time to look at the "ghost WU" issue, and put in the same "re-sync" code that Einstein uses. These ghosts _will_ occur, on every project - it's a given just because the internet isn't perfect. (Think of how many web pages give you errors and you have to refresh...) All anyone can do is fix things so that they get resent once they do occur.

There also is _much_ discussion over this past month on the "500"s among the developers, and one thing I have learned today from reading the archives of all of that, is that there is a section of the code that if an "unknown" error occurs, it falls through and says "500". So that makes the 500 worthless in determining exactly what happened. Hopefully the next BOINC version will be able to give a lot more info on the "real" errors, and then we'll be able to finally start fixing things instead of chasing our tails.

Nothing wrong with a queue size of 0.01 _or_ 0.5. :-)

Frustration understood - patience appreciated. Things _are_ improving.

---

I am getting no work from project today again, like others. Have one unit running on 1st system, 1 running and 3 qued on 2nd system.

If you run uFluids be sure to turn Windows updates off. MS sent a critical security update to my 1st system overnight, which caused it to reboot when mouse moved (srarting up the monitor). Had 12 hours (58%) completed on one of there units but with no chackpoints lost all. Back to start with the reboot.

Cheers
Ray

---

Pizza@Home Rays Place Rays place Forums

I am getting the odd "No Work...." but the next time it requests it seems to download ok, both my machines have not run out of Rosetta, it always seems to get one finally....

---

If you run uFluids ... no chackpoints lost all.

That is enough to disqualify uFluids for me. SZTAKI has also _removed_ the checkpoints they once had, from what I understand (I'd already left for other reasons).

I'm seeing problems at other projects that I think it's important for Rosetta to avoid. Time permitting, you've given me an idea for a new thread... I'll have to think on it.

---

Hang in there. We are going to expand our hardware soon. Is anyone running out of work?

Hang in there. We are going to expand our hardware soon. Is anyone running out of work?

Yep, I haven't been able to get any since yesterday and am down to about 3 hrs. work on one machine.

But I have LOTS of work for the Einstein@Home project. All I have to do is flip the switch, and who knows if and when it will be switched back.

F. Prefect

---

Hang in there. We are going to expand our hardware soon. Is anyone running out of work?

I just increased the amount of results to queue on my system... shouldn't be a problem now, changed the "connect to network every X days" from .1 to 1.0.

---

Well to minimize my no work from project i have each of my computers set to run 3 difrent projects (all run rosetta at 55%+ of the time). i am curently Deployed underway (US NAVY) so i have times when i can not check on my computers at all.. from my result updates it apears that 2 of my 4 computers are out or have run out of rosetta and are crunching other projects (corse could be that darn wwhatever you call the end of deadline swap to get it done stuff)

i definatly like rosetta and think it seems to be the best of the projects im running (see the stats on my sig) but i also have to hope i can get everything sorted out so before i get long term deployed (6 months) i can have everythign setup so that all 4 of my computers crunch effecently for the 6 months when i just have my wife at home who dosnt have the same drive i have to ensure the computers are crunching... actualy while writing this i also e-mailed my wife.. she was awake and confirmed 1 of the machines had no rosetta running on it..

I am definatly not going to falt you guys though.. since i am one of the people who came over from FaD and i know there are lots of us here and im sure alot more come friday when FaD closes. I and i hope the rest of us from FaD understand that we have the tendancy to overload a project when we all try grabing for new work units

just remember to be sure to keep us updated...
(OFFTOPIC)
I have noticed from what i understand that the results grapics from my understanding they havent been updated in 1 month (or is it my misinterpratation of them) i hate saying it but if you dont keep those updated (every few weeks at least) you are going to loose people to some projects that are more flashie

(/OFFTOPIC)

i hope the fixed you are talking about work and hold ... if they dont im afrade you will loose some users from this worthwhile project at least temporarly (im not leaveing or such but others may) curnetly i have discontunued crunching uFluids till they get there stuff better setup but i dont nessarly give up entirly on any project.

---

Hang in there. We are going to expand our hardware soon. Is anyone running out of work?

I will, and everyone else should, mark time. Not to be too inquisitive though, but how soon is "soon"? And what issues/problems do you expect the new installation to resolve or improve? Or are you just crossing fingers and hoping that bigger is better? The questions are somewhat rhetorical and tongue-in-cheek, so you needn't respond.

This no work from project also happens to me, and the frontpage says that there are 75k WU ready to be downloaded. My Rosetta-only machine is now idle and not doing any work (except for Azureus).

Admins should have made it so that I can get work with manual update, so that the message:-

12/15/2005 11:07:07 PM|rosetta@home|Note: not requesting new work or reporting results

did not appear as long as the cache is full.

---

Campeones everywhere!

Hey Guys,

cool down a little bit ;-)

• @belldandy_from_pleiades: As long as your client says: not requesting new work, you won't get any WU from the project; you have "a problem" locally with your client. The client thinks, he/she/it should do work for another project, you have attached. You can manually hit "update" as often as you want, the client will not ask for work at this project.
  
  To solve it, you can choose between
  
  
  
  • crunch the other project until it's time again for Rosetta,
    
  • change your ResourceShare
    
  • suspend the other attached project.
    

  
  
  

• All my 17 boxes got enough work from Rosetta in the last 24 hours, the only thing is, it needs a second hit, that usually comes with 4 - 5 Minutes
  
  
• Expanding there hardware will be a little more work; moving server-modules from one hardware to another; it could bring up greater problems for more than a few hours
  
  
• It will be important, that the project-team takes an intensiv look to the servers this Friday, the weekend and the coming week. I guess, they know this and will have this closed look.
  

---

Supporting BOINC, a great concept !

Admins should have made it so that I can get work with manual update

That's not something the projects can control. If your cache is full, and you owe a debt to another project, BOINC isn't going to let you get work that would either push the deadlines, or override the resource shares.

If your cache is "reasonable" (ie; not 10 days, but 4 or less) and you want more Rosetta work to be downloaded while you are there to watch it, simply "suspend" any other project(s), _then_ hit "update" on Rosetta. Suspend removes the other project's work from the calculation, so, for example, instead of "4 days of SETI and Rosetta", it suddenly wants "4 days of Rosetta".

Unless you get a "no work from project" message, that I understand they are working to fix, and is intermittent, you should see a flood of Rosetta work come in.

To test this, I just looked - I had two Rosetta results in my cache. I suspended my two other projects. Before I could even HIT "update" on Rosetta, it got four more results. (My cache is low, 0.25.)

EDIT:: Took too long, Yeti beat me to it. :-) Don't know how a giant furry thing can type that fast...

---

[EDIT:: Took too long, Yeti beat me to it. :-) Don't know how a giant furry thing can type that fast...

Maybe that is not fur, maybe they all fingers ...

---

EDIT:: Took too long, Yeti beat me to it. :-) Don't know how a giant furry thing can type that fast...

For a modern Yeti, carrying a smart-phone and a laptop through the snowy wood, it is important to have these speed while typing his comments; otherwise too many people could recognize him ...

LOL

---

Supporting BOINC, a great concept !

i definatly like rosetta and think it seems to be the best of the projects im running (see the stats on my sig) but i also have to hope i can get everything sorted out so before i get long term deployed (6 months) i can have everythign setup so that all 4 of my computers crunch effecently for the 6 months when i just have my wife at home who dosnt have the same drive i have to ensure the computers are crunching... actualy while writing this i also e-mailed my wife.. she was awake and confirmed 1 of the machines had no rosetta running on it..

From what I gather, you have some internet connectivity when deployed.

Did you consider making a few strateging pinholes in your firewall, and using VNC to check on the systems once a week or so.

If you're on a regular DSL / cable setup ( I don't know if you are) your IP address could change, but if so, DYNDNS can do a lot to help out.

Oh yeah, despite the good press it's getting, I'd suggest staying clear of TightVNC. One look at the processor usage of the server was enough for me - you do NOT want something like that running on a cruncher. :)

---