Well I thought a wheel had come off....but it looks like all 4 are off now although I get the odd upload and download still.

---

The data files may not get created properly when the work unit is turned into results. you can also have permission problems and a few other bug-a-boos that make the file go missing... including the server is so overloaded that the request does not get serviced by the driver and so the calling program times out and reports an error saying the file is missing... it is there, but the software did not find it in the allocated time.

---

I tried it with this WU. Let's see what it reports. I noticed there was a quick "upload failed" message in the transfer section. It's logical that it didn't report and will timeout. I'm thinking all you did was kill off your WUs and wasted all that crunch time.

I got curious also and aborted one of my results. I saw another thread where someone said to abort results to get them to upload. All it does is make them go away. This is the result I aborted: result
My result got a validate error and the other two are now waiting concensus. I think the goofy behavior is aborting results thinking that will solve the upload problem. I hope no one takes that advise.

With my WU in question I also got the VALIDATE ERROR. So obviously this scheme doesn't work and participants who "abort transfer" are just blindly throwing away their WUs.

In a nutshell: DON'T DO IT!

---

me@rescam.org

Boy the upload and download situation is just getting worse by the day. The backlog on my uploads keeps getting longer and longer. All fail about 20 seconds after start attempt. It definitly is traceable back to the long outage, when they fixed that breaker. I note that traffic seems to be constant from looking at the statistics for the Berkley link, so I doubt its cogent or UFO. But it is a rare wu that gets through. The last time we had this scenerio there was a problem with the firewall at Berkley, did someone forget to reset it when they brought the projects back up? The problem is on both the production site and the beta site. Been through long outages before, and am used to seeing the connections dropped message both here and on Classic, but usually it clears after a day or so, if my backlog is any indication and typical of others, it will take at least that to clear the backlog once they correct the problem. It does seem to be more on the upload side than the download side, as I am getting work, but I note that there are a large number of errored WU's due to download error. Since it is the weekend, I suspect this problem will last at least until Monday morning Berkley time.

---

Not likely the firewall ... I have been getting some through. Though, like you and many others, my backlog continues to climb. Especially as I am more successful downloading than up-loading ...

---

Not likely the firewall ... I have been getting some through. Though, like you and many others, my backlog continues to climb. Especially as I am more successful downloading than up-loading ...

Which is of course good, as long as you don't have workunits going passed their deadline. Fortunately that's not the case with my systems, so I'm happy being able to keep crunching while the folks at Berkeley are at it solving the upload issue.
Hope they do so before July 24th though.. :) That's when my first stuck wu's are going past their deadlines.

Greetings,
Speedy67

---

i haveno Probs here

9 SAH Wus are ready to send, but i also at

http://einstein.phys.uwm.edu/

in, and they have no pobs

Greetings from Germany NRW
Ulli

btw 9 WUs At SAH awe waiting for download at this time, at my Athlon64...

My suggestion to Berkeley: BE PROFESSIONAL!!!, i.e. add hardware, when it is needed, tell your "customers" what the problem is and tell it in time, again: TELL IT IN TIME!!! Everyone will understand (and accept) a problem. But what Berkeley does is nothing, at least nothing in coummunication to its crunchers (its customers!!!). I won't crunch any SETI wu anymore. I don't like this amateur behaviour and I will switch all idle CPU time to real science projects like LHC or Einstein or whatever.

---

My suggestion to Berkeley: BE PROFESSIONAL!!!, i.e. add hardware, when it is needed, tell your "customers" what the problem is and tell it in time, again: TELL IT IN TIME!!! Everyone will understand (and accept) a problem. But what Berkeley does is nothing, at least nothing in coummunication to its crunchers (its customers!!!). I won't crunch any SETI wu anymore. I don't like this amateur behaviour and I will switch all idle CPU time to real science projects like LHC or Einstein or whatever.

If you take a wee peek over at the Homepage, you will see, that Berkeley is doing just this: Telling us crunchers that there are problems.

July 15, 2005
The data server is continuing to drop connections, which is very unusual this long after an outage. We are working to figure out why.

As for the hardware:
They are no business, but a meager funded university lab. And we crunchers are no customers, as we don't pay them anything, but volunteers.

---

Gruesse vom Saenger

For questions about Boinc look in the BOINC-Wiki

Is this OS specific? My Windows 2K machine running BOINC 4.19 has 5 WU's seemingly unable to upload, but my slow Linux Box running BOINC 4.43 has no backlogs at all. The worst I've seen happen on it from the logs is that it will fail a first upload attempt, then try again 1 minute later and succeed.

---

<img src="http://boinc.mundayweb.com/one/stats.php?userID=531&amp;prj=1">

Is this OS specific? My Windows 2K machine running BOINC 4.19 has 5 WU's seemingly unable to upload, but my slow Linux Box running BOINC 4.43 has no backlogs at all. The worst I've seen happen on it from the logs is that it will fail a first upload attempt, then try again 1 minute later and succeed.

No i dont think that it&#180;s OS speciffic. The funny thing ist that all my machines @ home can upload just fine but all machines @work cannot upload at all.
That makes no sense at all. @ Work there are at least 150-200 WU&#180;s waiting for upload but i guess the work done by all machines is worthless because i think they will not be able to upload within the workunits deadline.

So all workunits are done for nothing.

---

Join BOINC United now!

My suggestion to Berkeley: BE PROFESSIONAL!!!, i.e. add hardware, when it is needed, tell your "customers" what the problem is and tell it in time, again: TELL IT IN TIME!!! Everyone will understand (and accept) a problem. But what Berkeley does is nothing, at least nothing in coummunication to its crunchers (its customers!!!). I won't crunch any SETI wu anymore. I don't like this amateur behaviour and I will switch all idle CPU time to real science projects like LHC or Einstein or whatever.

If you take a wee peek over at the Homepage, you will see, that Berkeley is doing just this: Telling us crunchers that there are problems.

July 15, 2005
The data server is continuing to drop connections, which is very unusual this long after an outage. We are working to figure out why.

As for the hardware:
They are no business, but a meager funded university lab. And we crunchers are no customers, as we don't pay them anything, but volunteers.

Naturally. But almost ONE WEEK LATER??? Don't worry, I read all the technical posts and all the stuff provided by Berkeley. But that is not the problem. I'm working in the IT business almost 20 years and I ACCEPT problems and the time needed to solve those problems. Nevertheless, if you are business or volunteer, you have to communicate with your customers (yes, even a volunteer is a customer). And if you are business and you can't supply your customers needs you are out. And if you run a project and you can't supply your projects needs, your project is out, or your projects fundings are not sufficient. Then you have to think about the sense of doing it on. Or you don't care about those, who help your project to proceed. In this way, even a voluntary project has "customers", and they should be treated as "customers".

---

I think you just told Seti to pull the plug.

If you have 20 years of experience in IT, you should be accustomed to the term "calm down, let it go". If it was Microsoft, you'd sag your shoulders, before lifting them again and dropping them again.

So why get all exited about this?

---

I think you just told Seti to pull the plug.

If you have 20 years of experience in IT, you should be accustomed to the term "calm down, let it go". If it was Microsoft, you'd sag your shoulders, before lifting them again and dropping them again.

So why get all exited about this?

I don't ask Seti to [b]pull the plug[/n]. I really want them to go on. But I ask them to go on "more professional". Look at this: A one day outage almost caused the project to go down, because of unsufficient ressources. Even today me, you and all the rest of our crunchers community can't up-/download, as it's supposed to work. There is still SETIClassic running. Account creation is disabled now, but why don't they close the whole old project and move all the spare hardware to the now running project SETIBoinc? What is the sense of doing two parallel projects, running the same science, when your fundings are limited???????? That's what I will not accept, and that is (for example) what I mean with the term "professional". By the way: I really like the whole SETI thing, I just want to get back what I'm offering, even when it is only some CPU cycles,

---

I think you have to keep repeating "Retry Now" under the "Transfers" tab, and under "Projects" tab "Update" it, so both are running at the same time. It just uploaded several of mine in a row. You just have to keep retrying and re updating (and it wouldn't hurt to try "File" and "Run Benchmarks" as well). I was able to upload about nine or so WUs for july 24th, doing this. Now I'm trying to upload July 26th (which only has 1 wu). Then I'm safe till july 27.

---

<a href="http://artapprentice.net/sumer/"><img src="http://artapprentice.net/sumer/sos.gif"></a>
<a href="http://artapprentice.net/sumer/">Secrets of Sumer</a>

Just closing down Classic will bring on a load of shit from the Classic crunchers that are still out there. So it has to be phased out. Next thing on the agenda will be the shutting down of the download server for Classic.

Then eventually when everyone has brought in their units, the upload server.

And only then will things here change. Maybe.
A seperate upload and download server for Seti@Boinc.

But that may go on for a while yet. Especially since we'll get a big outage again in August, to replace three more of them power breakers. So why rush now if you can wait?

---

I think you have to keep repeating "Retry Now" under the "Transfers" tab, and under "Projects" tab "Update" it, so both are running at the same time. It just uploaded several of mine in a row. You just have to keep retrying and re updating (and it wouldn't hurt to try "File" and "Run Benchmarks" as well). I was able to upload about nine or so WUs for july 24th, doing this. Now I'm trying to upload July 26th (which only has 1 wu). Then I'm safe till july 27.

That&#180;s fun.. 200 WU&#180;s to upload and just hitting retry now on all machines over the network...
And by the way there&#180;s no tab while running unix/linux or as a service...

But i keep searching for it somewhere at the shell .. LOL

---

Join BOINC United now!

(Javapersona puts on his old Cal sweatshirt in anticipation of his cheerleading routine).

A public university, meager funding, trying the best we can, all volunteers, blah blah blah. *sigh*

I am getting fed up with this. Each outage, which in an of itself takes a short time, is taking longer and longer to recover from. I am a little annoyed now because the backlog on my hosts mean I am attempting to contact SETI about every 3 to 5 minutes. Manually updating/retrying usually fails, too. To me, this is becoming a waste of bandwidth.

I agree with Tigher. End Classic Now! Leaving the "good hardware" there gives people less incentive to leave and is keeping BOINC SETI from achieving all that it could. The client programs are really painless installs now. There is really no reason, as far as I can tell, not to move over completely.

I am not going to go as far as Tigher and not accept new work, but I am going to scale back the computing share I give to SETI now. I had planned on doing this when some other BOINC projects came online. This quagmire has just accelerated my plans.

GO BEARS! MOVE ALL OF SETI HARDWARE TO BOINC!

---

I agree with Tigher. End Classic Now!

If they do it tomorrow, will you, Tigher and ecpa be around here 24 hours a day to ward off the complaints of everyone who was still crunching 5.00 to 600.00 units of Classic but have no where to return them to?

---

My suggestion to Berkeley: BE PROFESSIONAL!!!, i.e. add hardware, when it is needed, .

Add hardware with what? How do you propose they get this hardware to add? Do they have a magic genie that can grant them 3 wishes thus three brand new start of the art (or better) servers?

i must really be bored. I've been coming back to the seti forums. Maybe I am a masochist, or maybe I am just a gluton for punishment.

Jim

---