Since the last outage, I have a couple of hosts that refuse to recover. I started with 3 that would not download workunits at all. I was able to recover one by rerunning the Boinc installation (5.2.13), and selecting the repair option. On the other 2 systems, all are XP, that did not work. I tried removing 5.2.13, and installing 4.45, which had worked alright up to the point that I upgraded. Once reinstalled, the tools-options-http proxy setup was no longer available, it was greyed out. Once I reinstalled 5.2.13, the proxy options were available, but the error 500 on download attempts persisted. On one system, I downloaded and installed 5.2.15. The installation went without errors, but the error 500 has changed to "--- 1/10/2006 9:25:44 AM Network error: number of redirects hit maximum amount". This is only occuring on 2 of my 7 systems, so I don't see where the proxy server could be the issue. I would be interested to hear about any possible solutions. Thanks in advance.

---

An error code of 500 usually means a Berkley problem. Give it a couple more days and if it does fix itself write again.

---

Since the last outage, I have a couple of hosts that refuse to recover. I started with 3 that would not download workunits at all. I was able to recover one by rerunning the Boinc installation (5.2.13), and selecting the repair option. On the other 2 systems, all are XP, that did not work. I tried removing 5.2.13, and installing 4.45, which had worked alright up to the point that I upgraded. Once reinstalled, the tools-options-http proxy setup was no longer available, it was greyed out. Once I reinstalled 5.2.13, the proxy options were available, but the error 500 on download attempts persisted. On one system, I downloaded and installed 5.2.15. The installation went without errors, but the error 500 has changed to "--- 1/10/2006 9:25:44 AM Network error: number of redirects hit maximum amount". This is only occuring on 2 of my 7 systems, so I don't see where the proxy server could be the issue. I would be interested to hear about any possible solutions. Thanks in advance.

Maybe what I would like to know is, would anyone know how to remove Boinc so that a subsequent reinstall would be succesful? None of the systems will operate with the proxy server since I upgraded to 5.2.13 with a downgraded installation of 4.45, reinstalling 5.2.13, or 5.2.15. I just want to get on to crunching again like normal.

---

Maybe what I would like to know is, would anyone know how to remove Boinc so that a subsequent reinstall would be succesful?

the more garunteed way is to (preferaby finish all your work first by using "no new work") uninstall, delete the boinc folder, open "regedit.exe" and remove all references to BOINC, but be careful, changing stuff in the registry is dangerous, changing the wrong thing could kill windows

Maybe what I would like to know is, would anyone know how to remove Boinc so that a subsequent reinstall would be succesful?

the more garunteed way is to (preferaby finish all your work first by using "no new work") uninstall, delete the boinc folder, open "regedit.exe" and remove all references to BOINC, but be careful, changing stuff in the registry is dangerous, changing the wrong thing could kill windows

I have gone to Add-Remove and removed from there, followed by a reboot. I have deleted the entire directory. Upon reinstallation, 4.45 will no longer allow me to input an HTTP Proxy server. The menu window is there, but the selection is greyed out. If I put in a newer version of Boinc, 5.2.13 or 5.2.15, it automatically selects the correct proxy server settings. I went so far as to go to the registry and remove all instances of Boinc after a deinstall. Once reinstalled, I still get the error 500 / Redirects hit maximum amount message when I try to download workunits, the message varies with the exact version of Boinc installed. Four other systems connect properly to the project going through the same proxy server, all the systems had been working fine for months without any more problems than normal.

---

Maybe what I would like to know is, would anyone know how to remove Boinc so that a subsequent reinstall would be succesful?

the more garunteed way is to (preferaby finish all your work first by using "no new work") uninstall, delete the boinc folder, open "regedit.exe" and remove all references to BOINC, but be careful, changing stuff in the registry is dangerous, changing the wrong thing could kill windows

I have gone to Add-Remove and removed from there, followed by a reboot. I have deleted the entire directory. Upon reinstallation, 4.45 will no longer allow me to input an HTTP Proxy server. The menu window is there, but the selection is greyed out. If I put in a newer version of Boinc, 5.2.13 or 5.2.15, it automatically selects the correct proxy server settings. I went so far as to go to the registry and remove all instances of Boinc after a deinstall. Once reinstalled, I still get the error 500 / Redirects hit maximum amount message when I try to download workunits, the message varies with the exact version of Boinc installed. Four other systems connect properly to the project going through the same proxy server, all the systems had been working fine for months without any more problems than normal.

Just this afternoon, I got 2 workunits downloaded to one system, the rest all errored with a 500 error. Maybe this is coming from Berkeley?

---

Maybe what I would like to know is, would anyone know how to remove Boinc so that a subsequent reinstall would be succesful?

the more garunteed way is to (preferaby finish all your work first by using "no new work") uninstall, delete the boinc folder, open "regedit.exe" and remove all references to BOINC, but be careful, changing stuff in the registry is dangerous, changing the wrong thing could kill windows

I have gone to Add-Remove and removed from there, followed by a reboot. I have deleted the entire directory. Upon reinstallation, 4.45 will no longer allow me to input an HTTP Proxy server. The menu window is there, but the selection is greyed out. If I put in a newer version of Boinc, 5.2.13 or 5.2.15, it automatically selects the correct proxy server settings. I went so far as to go to the registry and remove all instances of Boinc after a deinstall. Once reinstalled, I still get the error 500 / Redirects hit maximum amount message when I try to download workunits, the message varies with the exact version of Boinc installed. Four other systems connect properly to the project going through the same proxy server, all the systems had been working fine for months without any more problems than normal.

Just this afternoon, I got 2 workunits downloaded to one system, the rest all errored with a 500 error. Maybe this is coming from Berkeley?

As of this morning, I have seen systems that did not have this previously, start showing it, and it is on both uploads and downloads. I think that people with persistant connections may not notice this, but on dial up, where you wait for the last transaction to finish before disconnecting, you see it because you really pay attention to the upload / download que in Boincview. With a broadband connection, you may not pay as much attention. Could this have something to do with overall connection speed? As 5 systems simultaneously attempt to download 2 workunits each on a 30k connection, you end up with one or two connections that get a speed of less than 100 bits per second. Maybe Berkeley's routers are trained to kill what may look like a DOS attack. Just a guess.

---

See if this thread helps:

"Interesting Post @ Einstein on upload failures"

---

Interesting Post @ Einstein on upload failures

Theres the link for it. :-)

Jeremy

---

Interesting Post @ Einstein on upload failures

Theres the link for it. :-)

Jeremy

Thanks both of you! I guess that that will be addressed in upcoming issues of Boinc. In the meantime, is the post suggesting that I should make the changes to the code, or just ride it out until the changes are implemented in a future release? (edit) (unedit)

---

Interesting Post @ Einstein on upload failures

Theres the link for it. :-)

Jeremy

Thanks both of you! I guess that that will be addressed in upcoming issues of Boinc. In the meantime, is the post suggesting that I should make the changes to the code, or just ride it out until the changes are implemented in a future release? (edit) (unedit)

Well if you want to change the code then fine but I guess its at your own risk. I would just hang on. If stuck try a near-by to CA proxy as it can help.

---

Interesting Post @ Einstein on upload failures

Theres the link for it. :-)

Jeremy

Thanks both of you! I guess that that will be addressed in upcoming issues of Boinc. In the meantime, is the post suggesting that I should make the changes to the code, or just ride it out until the changes are implemented in a future release? (edit) (unedit)

Well if you want to change the code then fine but I guess its at your own risk. I would just hang on. If stuck try a near-by to CA proxy as it can help.

"If stuck try a near-by to CA proxy as it can help." I don't know exactly what you mean by this. I guess: Someone has a proxy server for Berkeley? Would I change the URL for my connection rather than the one for Berkeley? ( http://setiathome.berkeley.edu )

---

Interesting Post @ Einstein on upload failures

Theres the link for it. :-)

Jeremy

Thanks both of you! I guess that that will be addressed in upcoming issues of Boinc. In the meantime, is the post suggesting that I should make the changes to the code, or just ride it out until the changes are implemented in a future release? (edit) (unedit)

Well if you want to change the code then fine but I guess its at your own risk. I would just hang on. If stuck try a near-by to CA proxy as it can help.

"If stuck try a near-by to CA proxy as it can help." I don't know exactly what you mean by this. I guess: Someone has a proxy server for Berkeley? Would I change the URL for my connection rather than the one for Berkeley? ( http://setiathome.berkeley.edu )

Well you open boinc manager and choose options. There is a tab for http proxy. You enter the proxy address and port there (plus password etc if needed) and away you go. From that point on boinc will connect to the proxy rather than to ucb - then the proxy passes it on to UCB. Its all about timers. If you cannot satisfy timers between you and ucb then you may be able to satisfy timers to a proxy and it may be able to satisy timers with ucb. You need to try a few though in my experience b4 u get 1 that works.

Here is a web page for some proxies. http://www.iantighe.com/boincers.htm . Try a few in or bear to CA and work back to your own location until you get one that works for you.

---

Interesting Post @ Einstein on upload failures

Theres the link for it. :-)

Jeremy

Thanks both of you! I guess that that will be addressed in upcoming issues of Boinc. In the meantime, is the post suggesting that I should make the changes to the code, or just ride it out until the changes are implemented in a future release? (edit) (unedit)

Well if you want to change the code then fine but I guess its at your own risk. I would just hang on. If stuck try a near-by to CA proxy as it can help.

"If stuck try a near-by to CA proxy as it can help." I don't know exactly what you mean by this. I guess: Someone has a proxy server for Berkeley? Would I change the URL for my connection rather than the one for Berkeley? ( http://setiathome.berkeley.edu )

Well you open boinc manager and choose options. There is a tab for http proxy. You enter the proxy address and port there (plus password etc if needed) and away you go. From that point on boinc will connect to the proxy rather than to ucb - then the proxy passes it on to UCB. Its all about timers. If you cannot satisfy timers between you and ucb then you may be able to satisfy timers to a proxy and it may be able to satisy timers with ucb. You need to try a few though in my experience b4 u get 1 that works.

Here is a web page for some proxies. http://www.iantighe.com/boincers.htm . Try a few in or bear to CA and work back to your own location until you get one that works for you.

I looked at your page, nice work! That was pretty much what I figured, a substitute for the standard url for hookups. I also found the Paul Buck Q&A pages, and it seems that the majority of the problems that I have could be coming from some packet corruption while enroute, timing problems, and the fact that UCB has done a lot of crunching data from hard drive failures and data base compressions and the like, there is an increased probability of specific file data corruption to occur. Now I guess I just have to wait for the data files to get old and die off?

---

Interesting Post @ Einstein on upload failures

Theres the link for it. :-)

Jeremy

Thanks both of you! I guess that that will be addressed in upcoming issues of Boinc. In the meantime, is the post suggesting that I should make the changes to the code, or just ride it out until the changes are implemented in a future release? (edit) (unedit)

Well if you want to change the code then fine but I guess its at your own risk. I would just hang on. If stuck try a near-by to CA proxy as it can help.

"If stuck try a near-by to CA proxy as it can help." I don't know exactly what you mean by this. I guess: Someone has a proxy server for Berkeley? Would I change the URL for my connection rather than the one for Berkeley? ( http://setiathome.berkeley.edu )

Well you open boinc manager and choose options. There is a tab for http proxy. You enter the proxy address and port there (plus password etc if needed) and away you go. From that point on boinc will connect to the proxy rather than to ucb - then the proxy passes it on to UCB. Its all about timers. If you cannot satisfy timers between you and ucb then you may be able to satisfy timers to a proxy and it may be able to satisy timers with ucb. You need to try a few though in my experience b4 u get 1 that works.

Here is a web page for some proxies. http://www.iantighe.com/boincers.htm . Try a few in or bear to CA and work back to your own location until you get one that works for you.

I looked at your page, nice work! That was pretty much what I figured, a substitute for the standard url for hookups. I also found the Paul Buck Q&A pages, and it seems that the majority of the problems that I have could be coming from some packet corruption while enroute, timing problems, and the fact that UCB has done a lot of crunching data from hard drive failures and data base compressions and the like, there is an increased probability of specific file data corruption to occur. Now I guess I just have to wait for the data files to get old and die off?

I normally wouldn't do this, but I attached 2 of the systems to Einstein, and there is no problem getting work, and the application files etc. I can only conclude that Berkeley has some kind of issue that I can't resolve from this end. I thank those that gave information to help me with this, a lot of what is misunderstood I believe, is that the software requires a lot of patience, while it tries to correct issues by itself. With Classic, there were a number of things that could be done to affect an issue, and the results were by far more immediate.

---

I normally wouldn't do this, but I attached 2 of the systems to Einstein, and there is no problem getting work, and the application files etc. I can only conclude that Berkeley has some kind of issue that I can't resolve from this end. I thank those that gave information to help me with this, a lot of what is misunderstood I believe, is that the software requires a lot of patience, while it tries to correct issues by itself. With Classic, there were a number of things that could be done to affect an issue, and the results were by far more immediate.

So, the question becomes "what is the difference between SETI and Einstein?"

... and also, what isn't (BOINC handles all communications, for both projects).

The problem is that your problem is not well understood. Most of us aren't having trouble. Those who do usually have some projects that work without error, and others that are really difficult.

---

I normally wouldn't do this, but I attached 2 of the systems to Einstein, and there is no problem getting work, and the application files etc. I can only conclude that Berkeley has some kind of issue that I can't resolve from this end. I thank those that gave information to help me with this, a lot of what is misunderstood I believe, is that the software requires a lot of patience, while it tries to correct issues by itself. With Classic, there were a number of things that could be done to affect an issue, and the results were by far more immediate.

So, the question becomes "what is the difference between SETI and Einstein?"

... and also, what isn't (BOINC handles all communications, for both projects).

The problem is that your problem is not well understood. Most of us aren't having trouble. Those who do usually have some projects that work without error, and others that are really difficult.

And remeber the patch for this came from CPDN for a user having....yes 500 problems.!!. Its generic...seti sees it most because of numbers and volume.

---

And remeber the patch for this came from CPDN for a user having....yes 500 problems.!!. Its generic...seti sees it most because of numbers and volume.

Yes, but...

What is "it" anyway? I mean, I know "it" is 500 errors, and I know that 500 errors come from web servers. I know that some things inside BOINC report 500 errors, and I know most users aren't having problems.

So, if we're going to solve "it" we're going to have to figure out why someone like Leo has the problem, and someone else doesn't.

But then I know you knew that, Ian....

-- Ned

---

And remeber the patch for this came from CPDN for a user having....yes 500 problems.!!. Its generic...seti sees it most because of numbers and volume.

Yes, but...

What is "it" anyway? I mean, I know "it" is 500 errors, and I know that 500 errors come from web servers. I know that some things inside BOINC report 500 errors, and I know most users aren't having problems.

So, if we're going to solve "it" we're going to have to figure out why someone like Leo has the problem, and someone else doesn't.

But then I know you knew that, Ian....

-- Ned

Might be best to start with what I don't know. I don't know how many separate issues we have here. I don't know how many things report 500 as an error. I don't know if its fastcgi or apache or server or boinc cc or user LAN firmware. Actually I know nothing to very little for certain. Hmmm makes me wonder why I do this lol!

We know one thing. Thyme Lawn saw, and I agree, that the scheduler replies to a request and sometimes it sends a HTTP 200 OK and things go well and sometimes it sends a 100 CONTINUE and the boinc client does not process it....in fact it closes the socket and says thats an error mate here have this 500. So we need to get that clear first then we might see some solved and some not still. But remember when I started out with 500 last July 5% of all transactions at the seti server end finished that way...not now its 10% lol not really!!!! Its actually 0% at the server so its being narrowed down. We make prgress slowly here I feel.

Well I can say Rom sent me a 5.3.11 (Dr A put the fix in .11) and I got a fella to try it. Damn it made no difference and I wait for Rom to confirm it was indeed 5.3.11 as it announced itself as 5.3.10. So we wait further. I am intrigued though that elsewhere a firmware revision for Linksys users can help and for some MTU size changes can help ....I bet we got 10 different errors in this lot!

---

And remeber the patch for this came from CPDN for a user having....yes 500 problems.!!. Its generic...seti sees it most because of numbers and volume.

Yes, but...

What is "it" anyway? I mean, I know "it" is 500 errors, and I know that 500 errors come from web servers. I know that some things inside BOINC report 500 errors, and I know most users aren't having problems.

So, if we're going to solve "it" we're going to have to figure out why someone like Leo has the problem, and someone else doesn't.

But then I know you knew that, Ian....

-- Ned

Might be best to start with what I don't know. I don't know how many separate issues we have here. I don't know how many things report 500 as an error. I don't know if its fastcgi or apache or server or boinc cc or user LAN firmware. Actually I know nothing to very little for certain. Hmmm makes me wonder why I do this lol!

We know one thing. Thyme Lawn saw, and I agree, that the scheduler replies to a request and sometimes it sends a HTTP 200 OK and things go well and sometimes it sends a 100 CONTINUE and the boinc client does not process it....in fact it closes the socket and says thats an error mate here have this 500. So we need to get that clear first then we might see some solved and some not still. But remember when I started out with 500 last July 5% of all transactions at the seti server end finished that way...not now its 10% lol not really!!!! Its actually 0% at the server so its being narrowed down. We make prgress slowly here I feel.

Well I can say Rom sent me a 5.3.11 (Dr A put the fix in .11) and I got a fella to try it. Damn it made no difference and I wait for Rom to confirm it was indeed 5.3.11 as it announced itself as 5.3.10. So we wait further. I am intrigued though that elsewhere a firmware revision for Linksys users can help and for some MTU size changes can help ....I bet we got 10 different errors in this lot!

The whole Linksys discussion below is interesting. The idea that there is an uninitialized variable explains alot.

My connection supports MTU=1500 and it damn well better, I pay enough for it. For $15/month, PPPoE is fine, and I'd be okay with 1492 or lower.

The "100 continue" issue sounds very real. If the servers are going to send that, the client should support it, end of story.

... and the "500" error should not be reported by the BOINC client. I think that's fixed in the current alpha.

-- Ned

---