Stats are back! Can anyone upload?

[HaloJones]

I can see the stats pages. Can anyone verify that uploads are on again?

[mighty]

I managed to upload 5 generations and then it stopped. I have not used the -v swtich yet, but it seems the default value is rather high. It has been waiting for the response for generation 5 for alomst 10 minutes now...

Disregard the above. I was using the -ut switch, so -v does not apply. My bad.

[pfb]

Seems to be uploading OK...

[FoBoT]

i am getting a few gens. through on a few pc's
i am sure it is swamped
better to just wait a bit i think

[Ripper [USD]]

I can upload, but the uploads are limited (or appear to be limited) to 5 buffered generations per pass.

It will take HOURS, of not days, to upload all my buffered results (over 1000).

[tpdooley]

On one of my machines that would have 277 buffered gens if the servers were still down, I've only got around 230 buffered gens now. As soon as some of the online folders have uploaded all their buffered gens, the upload speed will start improving. And since this is uploading 5 times faster than they're generating gens, it should be finished uploading all the online folders by tomorrow.

[HaloJones]

I'll let you know when I've successfully uploaded the 8000+ gens I have on my uploader from last Friday :sleepy: I'm not even thinking about the 20000+ sitting on the machines themselves!

[pfb]

Just be glad it wasn't a quick protein this happened on

[erk]

Originally posted by pfb
Just be glad it wasn't a quick protein this happened on

Which raises an interesting issue, the last protein was twice as fast as this one, if we get another short one like that again can the ticketing system/server keep up? It would be interesting to know how many hours the server outage was for, and how many hours it takes to clear the backlog.

At the moment most of my online machines are not even half way through uploading their cached gens, and I have several offline machines to go. If they take more than about 36 hours to clear the backlog then there may be serious trouble on short proteins.

[HaloJones]

Originally posted by HaloJones
I'll let you know when I've successfully uploaded the 8000+ gens I have on my uploader from last Friday :sleepy: I'm not even thinking about the 20000+ sitting on the machines themselves!

Still uploading after 24 hours. Now have 25000+ sitting on the off-network machines themselves. Hope this backlog clears soon!

Oh and I just passed 2 billion for the whole DF thing

[deranged128[OCAU]]

My backlog is clearing gradually with still around 5,000+ gens cached.

An early prediction for a short/faster protein will be the return of 200 structures per generaion (or possibly some other number like 150) to contain the amount of uploads to a manageable level within the current system.

I think like many I was surprised that the system was left offline for as long as it was. I'm well aware that it was a Public Holiday over there on the Monday but feel with the amount of resources that are being contributed to this project from around the world that looking at the problem within 24 hours of it occuring should be a necessity.

On another note, I certainly hope the next changeover happens no later than Tuesday of whatever week, or if not possible, put back a week, rather than the problems from last time when it was left too late in the week.

[Grumpy]

We need 2 weeks to clear the backlog then a eweek to rest

[Ripper [USD]]

Well, after watching the numbers grow and the backlog getting cleared out, i noticed that over half of the generations are still on the system. I currently have 937 generations that still need to be uploaded.

I restarted each client that has generations buffered and noticed that each of them states that it cannot find the previous generation and then goes back to doing work. So, somewhere in this MESSED UP SYSTEM, I have done a lot of work but can't get credit for the work done, as it will never upload the 937 generations.

I tried to contact the "owners" of the project and offered to let them use my server (Linux based) for free, to which they replied that they don't need the help, as the problems are temporary.

I might believe that the problem is TEMPORARY, except that it happens so often. I don't have a problem with putting my computers on jobs that are worthwhile, but I don't need to spend HUNDREDS of dollars on the electricty and then get no credit for it. Credit is the ONLY thing that we get other than the satisfaction that we are helping. Well, it is NO HELP if I can't upload the results after crunching for DAYS only to have a system that is so unreliable throw out all my results cause they can't keep people out of the systems.

I QUIT.

[Crabbleboys]

Originally posted by Ripper [USD]


I might believe that the problem is TEMPORARY, except that it happens so often . I don't have a problem with putting my computers on jobs that are worthwhile, but I don't need to spend HUNDREDS of dollars on the electricty and

I QUIT.

And that is the real problem. Even without the error that I have "cant find missing file "handle" etc it is a bloody pain and as I have stated I have only a 56k (41/44k)connection. The time involved uploading since the ticketing was brought in is a real problem even with a large work unit. Had this happened on a small unit i dread to think of the time involved trying to upload all the results.

Without wishing to offend it sometimes seems that this project is being run on a shoestring budget. I do have a real life and I very much enjoy it and nursing any project loses it's appeal eventually.

[Grumpy]

These files that go missing must be somewhere, I looked in my washing machine, but only found a red and blue sock

This problem has been around a very long time with no sign of a fix, nor a reason why it happens

[Ripper [USD]]

Well, and the fact that the Admin believes that the problems are only temporary, like this is something that just happened this one time, makes me wonder what all this is for.

They throw this problem off so casually, I think that the problem is NOT the server but the poor attitude of the staff running this project.

I have removed all my machines off this project and won't come back to this project to do anything, ever.

I don't like the attitude of those running the project or non-serious manner in which they handle complete loss of data. Hey, they don't pay the power bills and so what do they care?

I am finding that many VOLUNTEER organizations want you to pay for everything, but as for what you actually do within the organization or what happens to you while you are helping is of little consequence to them.

I stopped working as a Deputy Sheriff because of this, recently resigned from Boy Scouts because of this, I guess that DCing is next.

If you want my help, then you have to be as serious about the project you want us to help you with, as those of us working in it.

Does this give you any clues about my attitude towards Distributed Folding???