After this long outage, I got d/l's back yesturday but no uploads, I sat this morning and kept hitting the retry transfer button for the WU's waiting for U/L and just got them to go through - its obvious that the server can not take all of the U/L from everyone at the same time, so when the next 2 WU finish Crunching I'm guessing the problem will be the same, until the Ready to Send gets ALOT lower (250,000+ Computers Ready to Send currently).
The problem now is the Waiting for Validation has gone up again on the Server Stats Page. Are we going to see another long outage while those systems have to catch up again?




Kai

---

Please go back and read the Technical News, it is right above the Server Status link. It will answer all your questions.

---

Hmm, no problems on D/L or U/L. Maybe because I just let it do it's own thing instead of hammering the Update button.

-Mr. anon

---

This is the best information that anyone can give.
LET IT DO ITs OWN THING, and it will eventually clear out.
I'm sure 90% of the users now are at the point where they are frustrated and want to try to push theirs through so that theirs gets counted,
BUT
Each time each of us hit the retry buttons, you add unnecessary traffic, and that is one reason why so many can't get through.
There are even some who have modified it so that their retries happen quicker than the normal boinc timeouts... shame on you. All that does is bog down the server with multiple unnecessary hits and therefore FAILS.

Give it a break, let it do its own thing, and everyone will be happier in a day or so.

Patience, Persistence, Truth

Hmm, no problems on D/L or U/L. Maybe because I just let it do it's own thing instead of hammering the Update button.

-Mr. anon

---

Thanks, and Keep on crunchin'
John Henry KI4JPL
Sevierville TN

I started with nothing,
and I still have some of it left.
<img src="http://www.boincstats.com/stats/banner.php?cpid=989478996ebd8eadba8f0809051cdde2">

After this long outage, I got d/l's back yesturday but no uploads, I sat this morning and kept hitting the retry transfer button for the WU's waiting for U/L and just got them to go through - its obvious that the server can not take all of the U/L from everyone at the same time, so when the next 2 WU finish Crunching I'm guessing the problem will be the same, until the Ready to Send gets ALOT lower (250,000+ Computers Ready to Send currently).
The problem now is the Waiting for Validation has gone up again on the Server Stats Page. Are we going to see another long outage while those systems have to catch up again?




Kai

the ready to send on the status page is the number of WU ready to be sent out not the number of WU ready to send back.

---

This is the best information that anyone can give.
LET IT DO ITs OWN THING, and it will eventually clear out.
I'm sure 90% of the users now are at the point where they are frustrated and want to try to push theirs through so that theirs gets counted,
BUT
Each time each of us hit the retry buttons, you add unnecessary traffic, and that is one reason why so many can't get through.
There are even some who have modified it so that their retries happen quicker than the normal boinc timeouts... shame on you. All that does is bog down the server with multiple unnecessary hits and therefore FAILS.

Give it a break, let it do its own thing, and everyone will be happier in a day or so.

Patience, Persistence, Truth

Hmm, no problems on D/L or U/L. Maybe because I just let it do it's own thing instead of hammering the Update button.

-Mr. anon

When I tried to push an upload, it failed. When I just left it alone to upload whenever the client felt like it, it worked.

So let me repeat what the others said:
Let the software do its thing, and everything will clear up!

---

"Are you suggesting coconuts migrate?"

There are even some who have modified it so that their retries happen quicker than the normal boinc timeouts...

Out of interest, how do you know this?

---

There are even some who have modified it so that their retries happen quicker than the normal boinc timeouts...

Out of interest, how do you know this?

Out of interest, do you really think it's a good idea to post this info in a public forum?

;-)

---

"Are you suggesting coconuts migrate?"

There are even some who have modified it so that their retries happen quicker than the normal boinc timeouts...

Out of interest, how do you know this?

Out of interest, do you really think it's a good idea to post this info in a public forum?;-)

NO it is NOT! BUT if you have not already done so you could post it in the developers email list
List-Archive: http://ssl.berkeley.edu/pipermail/boinc_projects
List-Post: mailto:boinc_projects@ssl.berkeley.edu
List-Help: mailto:boinc_projects-request@ssl.berkeley.edu?subject=help
List-Subscribe: http://www.ssl.berkeley.edu/mailman/listinfo/boinc_projects

---

There are even some who have modified it so that their retries happen quicker than the normal boinc timeouts...

Out of interest, how do you know this?

Out of interest, do you really think it's a good idea to post this info in a public forum?;-)

NO it is NOT! BUT if you have not already done so you could post it in the developers email list
List-Archive: http://ssl.berkeley.edu/pipermail/boinc_projects
List-Post: mailto:boinc_projects@ssl.berkeley.edu
List-Help: mailto:boinc_projects-request@ssl.berkeley.edu?subject=help
List-Subscribe: http://www.ssl.berkeley.edu/mailman/listinfo/boinc_projects

I wasn't interested in how it was done but how he knew it was done.

And yes I do know that this will create an interest in how it was done and no I don't think he should supply the info.

So:

Out of interest, how do you know this?

---

Frankly, I think that there is a problem with the U/L D/L retry code.

When the systems first came back up, all of my WUs in 'ready to report' status uploaded almost immediately. As new WU's complete, they upload. None of my WUs that are stuck in 'retry' status have ever uploaded. Eventually, after some number of retries, they will revert to 'ready to report' and then they upload. Downloads in 'retry' status were not completing and blocking additional attempts. When they finally abort, a new D/L is started and completes.

This does not seem to be a 'contention' problem, because the chances of a single 'ready to report' W/U making it in a single attempt while 30 or so 'retry' WU's cannot after several hundred attempts over several days seems unlikely.

I have noticed similar behaviour after previous outages.

---

...don

Frankly, I think that there is a problem with the U/L D/L retry code.

When the systems first came back up, all of my WUs in 'ready to report' status uploaded almost immediately. As new WU's complete, they upload. None of my WUs that are stuck in 'retry' status have ever uploaded. Eventually, after some number of retries, they will revert to 'ready to report' and then they upload. Downloads in 'retry' status were not completing and blocking additional attempts. When they finally abort, a new D/L is started and completes.

This does not seem to be a 'contention' problem, because the chances of a single 'ready to report' W/U making it in a single attempt while 30 or so 'retry' WU's cannot after several hundred attempts over several days seems unlikely.

I have noticed similar behaviour after previous outages.

From the "Technical News": "Well, we uncovered another problem - the upload/download server was so busy it randomly lost NFS mounts, including necessary things like /usr/local. So the file_upload_handler was flailing throughout the course of the evening. This morning (after the usual Wednesday database backup outage) we determined this was an automounter problem, put in some hard mounts for required partitions, and so far it's been working pretty well (though still very far from catching up. We're dropping hundreds of connections per second - only a lucky 20-30 RPCs/sec are getting through)."

Frankly, I think that there is a problem with the U/L D/L retry code.

When the systems first came back up, all of my WUs in 'ready to report' status uploaded almost immediately. As new WU's complete, they upload. None of my WUs that are stuck in 'retry' status have ever uploaded. Eventually, after some number of retries, they will revert to 'ready to report' and then they upload. Downloads in 'retry' status were not completing and blocking additional attempts. When they finally abort, a new D/L is started and completes.

This does not seem to be a 'contention' problem, because the chances of a single 'ready to report' W/U making it in a single attempt while 30 or so 'retry' WU's cannot after several hundred attempts over several days seems unlikely.

I have noticed similar behaviour after previous outages.

From the "Technical News": "Well, we uncovered another problem - the upload/download server was so busy it randomly lost NFS mounts, including necessary things like /usr/local. So the file_upload_handler was flailing throughout the course of the evening. This morning (after the usual Wednesday database backup outage) we determined this was an automounter problem, put in some hard mounts for required partitions, and so far it's been working pretty well (though still very far from catching up. We're dropping hundreds of connections per second - only a lucky 20-30 RPCs/sec are getting through)."

That doesn't seem to be relevant to his particular experiences.

---