Just a thought that maybe this is a planned outage. Think about it. The last 2 or 3 outages all began on a Friday night and continued until Monday morning. Could it be that the developers need data on how Boinc behaves after an extended outage and therefore create an outage? On the Seti Classic site data is still flowing in and out and Classic uses the same Cogent ssl link for upload and download. The Cogent bandwidth charts show the drop exactly when Boinc went down and the reduced bandwidth still being used by Classic. I don’t believe there is any IO occurring for Boinc/Seti.

It is also strange that the Server Status page shows that there are 5,395 WU waiting for validation and 4 validators running. I have never seen the value this high except AFTER an extended outage.

Since Boinc and Seti Classic have NEVER admitted to a software problem causing an outage I would expect that they will announce some sort of hardware failure on Monday morning. Then after the outage is repaired (router turned back on?) we will all see how the current version of Boinc responds to an extended outage and the flood of IO requests that follows.

Anyway if this is all true then the developers have the right to test their baby under all conditions to make sure it is performing as expected and indeed I expect these sort of tests. They are in the process of creating a very complex piece of software that must perform under countless user environments and perform flawlessly. NOT an easy task. These guys and gals certainly have my respect and I wish them success. I wish them success because if they succeed then I have more fun in my distributed computing hobby.

It is after all a hobby isn’t it? Well it should be!

---

Boinc....Boinc....Boinc....Boinc....

WOw, I like the theory.

I wonder if they'll also test what happens when they keep it functioning at 100% peak efficiency 24/7 for more then a month :)

---

Join team BOINCstats

I think bad things tend to happen during the weekends because less people are paying attention to the servers and therefore not fixing problems right away (though I did spend two hours last night screwing around with the web server which was all hung up).

But in this case, I think there is a bunch of minor problems caused by the recent web domain switch which in turn caused more demand on the whole set of BOINC servers. Basically this is yet another BOINC stress test. One odd thing is that it is too difficult to get a news item posted on the front page (since the machine hosting the pages is having NFS issues I don't want to deal with right now).

And it is the weekend, and I have plans, so.. see you later.

- Matt

---

-- BOINC/SETI@home network/web/science/development person
-- "Any idiot can have a good idea. What is hard is to do it." - Jeanne-Claude

Now that all my "slugs" are up to 4.44, it's just what I was waiting for to see how long and the specific actions BOINC will take for them to get back to steady state.

Although it will have to last about another 20 hours or so for it to have any effect on them except for filling the log with "This Ride is Closed" messages from SETI when they try to report.

Alinator

Thanks for the "head's up" there Matt. Guess it's hunker down and crunch like hell till Monday ...

Anyway if this is all true then the developers have the right to test their baby under all conditions to make sure it is performing as expected and indeed I expect these sort of tests. They are in the process of creating a very complex piece of software that must perform under countless user environments and perform flawlessly. NOT an easy task. These guys and gals certainly have my respect and I wish them success. I wish them success because if they succeed then I have more fun in my distributed computing hobby.

Don't forget that BOINC is supposed to be pretty resilient under these conditions: things like redundant connections are not strictly required.

.... and frankly, if we see a few more outages and nobody gets hurt, well, maybe a few people will be less prone to panic attacks every time there is some little hiccup.

I've got work from four different projects, and it looks like at least two are running perfectly, with much work to do. I'm happy.

---

I think bad things tend to happen during the weekends because less people are paying attention to the servers and therefore not fixing problems right away (though I did spend two hours last night screwing around with the web server which was all hung up).

But in this case, I think there is a bunch of minor problems caused by the recent web domain switch which in turn caused more demand on the whole set of BOINC servers. Basically this is yet another BOINC stress test. One odd thing is that it is too difficult to get a news item posted on the front page (since the machine hosting the pages is having NFS issues I don't want to deal with right now).

And it is the weekend, and I have plans, so.. see you later.

- Matt

Thanks for your Answer, and also for your good Work here,

Have i nice Weekend

i have an Cache of 3 Days, hope that Monday all is going the right way.

Greetings from Germany NRW
Ulli S@h Berkeley's Staff Friends Club m7 ©

[font='fixedsys,courier']I like the theory, but the real conspiracy is the Breakers' UotD...[/font]

---

[font='fixedsys,courier']I like the theory, but the real conspiracy is the Breakers' UotD...[/font]

But that's just for the forums, anyone know of a conspiracy to break anything else BOINC?

:)

---

It is a sad sad day when someone takes your spoon away from you...

I think bad things tend to happen during the weekends because less people are paying attention to the servers and therefore not fixing problems right away (though I did spend two hours last night screwing around with the web server which was all hung up).

But in this case, I think there is a bunch of minor problems caused by the recent web domain switch which in turn caused more demand on the whole set of BOINC servers. Basically this is yet another BOINC stress test. One odd thing is that it is too difficult to get a news item posted on the front page (since the machine hosting the pages is having NFS issues I don't want to deal with right now).

And it is the weekend, and I have plans, so.. see you later.

- Matt

Matt have a great weekend and thanks for the insight on the weekend issue.

---

Just a thought that maybe this is a planned outage. Think about it. The last 2 or 3 outages all began on a Friday night and continued until Monday morning. Could it be that the developers need data on how Boinc behaves after an extended outage and therefore create an outage? On the Seti Classic site data is still flowing in and out and Classic uses the same Cogent ssl link for upload and download. The Cogent bandwidth charts show the drop exactly when Boinc went down and the reduced bandwidth still being used by Classic. I don’t believe there is any IO occurring for Boinc/Seti.

It is also strange that the Server Status page shows that there are 5,395 WU waiting for validation and 4 validators running. I have never seen the value this high except AFTER an extended outage.

Since Boinc and Seti Classic have NEVER admitted to a software problem causing an outage I would expect that they will announce some sort of hardware failure on Monday morning. Then after the outage is repaired (router turned back on?) we will all see how the current version of Boinc responds to an extended outage and the flood of IO requests that follows.

Anyway if this is all true then the developers have the right to test their baby under all conditions to make sure it is performing as expected and indeed I expect these sort of tests. They are in the process of creating a very complex piece of software that must perform under countless user environments and perform flawlessly. NOT an easy task. These guys and gals certainly have my respect and I wish them success. I wish them success because if they succeed then I have more fun in my distributed computing hobby.

It is after all a hobby isn’t it? Well it should be!

Hmmmmmmmmm

I need to upload 6 wu's and download 2 wu's.. Scheduler finally said I need "something" and spit 4 new workunits... Somehow I fear with the unscheduled "outage" the six completed are "lost!" and the two downloads are to be reported as "Client Error!" Ohh it was a glitch!

But then what do you expect from "Amateurs"

I suspect that David is out trying to sell waterfront property in Florida...

Al

But then what do you expect from "Amateurs"

I suspect that David is out trying to sell waterfront property in Florida...

Al

Why the ad-hominum attack? Why call someone unskilled or a fraud?

If you don't like some technical aspect, then fine, argue the technical merits. There are plenty here who will disagree with you, and from the discussion we can maybe figure out how to make BOINC better.

... but the ad-hominum arguments are beyond the valley of good taste.

If you have a technical argument, make it. If there is any merit to your argument, name calling won't be needed.

---

Conspiracy Theory

But then what do you expect from "Amateurs"

I suspect that David is out trying to sell waterfront property in Florida...

Al

Why the ad-hominum attack? Why call someone unskilled or a fraud?

If you don't like some technical aspect, then fine, argue the technical merits. There are plenty here who will disagree with you, and from the discussion we can maybe figure out how to make BOINC better.

... but the ad-hominum arguments are beyond the valley of good taste.

If you have a technical argument, make it. If there is any merit to your argument, name calling won't be needed.

Maybe, since I have donated unused CPU cycles of 14 computers in my home computer lab "ages ago" and have seen most of the things that "might have been" avoided. Several hundred thousand computer hours later, am a bit tired of the "Same Mistakes"... Those mistakes tend to display that his emphasys is not on the "ultimate stability" of what the project he founded. Maybe his emphasys is on National Press, when he can grab it. I have never had an issue with Einstein, I have had issues with Seti Boinc and Seti Classic... In that I am not alone. The Forum is littered with "technical" proof... If this tires you go read Seti Classic...

Look at what "should be posted" in the area that is set aside to communicate "issues" that affect users that are donating their CPU cycles... Oops Sorry that would presume "professionalism." Look at how many times past issues were ignored and the Forums were littered with "whinning." Look at how many volunteers with leaked information, spend countless hours trying to soothe the troubled masses...

How easy is it for a "New User" to get things going? Paul Buck has spent countless hours writing about how, what and where... You still keep seeing the same questions... Why? Because the client is not ready for the "average internet user." With the Gigabit link going down (unresponded) on multiple occasions. Hmmmm no redunacy in conecectitvy (even a redundant T3/DS3 would be a help on ocasions). Various server issues that should not have happened. The power outages hmmmm allk this in only in the last two months... I could go search the Forum for more...

Then the Credit issue, create a simple benchmark workunit. When You complete it you have credits for the CPU time to complete the work unit. It does not matter if it takes two hours or two days, you receive credit for the CPU cycles you donate. Then scheduler can attempt to match machines to so that those that take two days will be matched with those tht take two days to complete a workunit.

So yes part of the proof is in what is Not said that should be, by the offical "Seti." Which would avert having to create a place like this so people can get things off their chest, find people that can help them get started (because the client is not user friendly) and giving people like Paul Buck someting to do (BTW Thank You Paul, your guides are quite comprehensive {Tip of the Hat})...

So with all that said, what makes this project so professional?

I am only here because I believe in the what Seti stands for. I reserve to right to look at how it runs and within my knowledge of networks, system administration and state this is one of the "poorest" run networks in the world. I know of small offices ran by secretary's that "have" better uptime than Seti "has" or "will ever achieve." I work with a slightly larger network that with all its complexity does not have "similar" issues.

So yes, I believe that is "now" down to name calling...

Yes I would look forward to David's reply!

---

Conspiracy Theory

But then what do you expect from "Amateurs"

I suspect that David is out trying to sell waterfront property in Florida...

Al

Why the ad-hominum attack? Why call someone unskilled or a fraud?

If you don't like some technical aspect, then fine, argue the technical merits. There are plenty here who will disagree with you, and from the discussion we can maybe figure out how to make BOINC better.

... but the ad-hominum arguments are beyond the valley of good taste.

If you have a technical argument, make it. If there is any merit to your argument, name calling won't be needed.

Maybe, since I have donated unused CPU cycles of 14 computers in my home computer lab "ages ago" and have seen most of the things that "might have been" avoided. Several hundred thousand computer hours later, am a bit tired of the "Same Mistakes"... Those mistakes tend to display that his emphasys is not on the "ultimate stability" of what the project he founded. Maybe his emphasys is on National Press, when he can grab it. I have never had an issue with Einstein, I have had issues with Seti Boinc and Seti Classic... In that I am not alone. The Forum is littered with "technical" proof... If this tires you go read Seti Classic...

Look at what "should be posted" in the area that is set aside to communicate "issues" that affect users that are donating their CPU cycles... Oops Sorry that would presume "professionalism." Look at how many times past issues were ignored and the Forums were littered with "whinning." Look at how many volunteers with leaked information, spend countless hours trying to soothe the troubled masses...

How easy is it for a "New User" to get things going? Paul Buck has spent countless hours writing about how, what and where... You still keep seeing the same questions... Why? Because the client is not ready for the "average internet user." With the Gigabit link going down (unresponded) on multiple occasions. Hmmmm no redunacy in conecectitvy (even a redundant T3/DS3 would be a help on ocasions). Various server issues that should not have happened. The power outages hmmmm allk this in only in the last two months... I could go search the Forum for more...

Then the Credit issue, create a simple benchmark workunit. When You complete it you have credits for the CPU time to complete the work unit. It does not matter if it takes two hours or two days, you receive credit for the CPU cycles you donate. Then scheduler can attempt to match machines to so that those that take two days will be matched with those tht take two days to complete a workunit.

So yes part of the proof is in what is Not said that should be, by the offical "Seti." Which would avert having to create a place like this so people can get things off their chest, find people that can help them get started (because the client is not user friendly) and giving people like Paul Buck someting to do (BTW Thank You Paul, your guides are quite comprehensive {Tip of the Hat})...

So with all that said, what makes this project so professional?

I am only here because I believe in the what Seti stands for. I reserve to right to look at how it runs and within my knowledge of networks, system administration and state this is one of the "poorest" run networks in the world. I know of small offices ran by secretary's that "have" better uptime than Seti "has" or "will ever achieve." I work with a slightly larger network that with all its complexity does not have "similar" issues.

So yes, I believe that is "now" down to name calling...

Yes I would look forward to David's reply!

Certainly the name calling and harrassment will inspire him to respond. Good idea.

You asked for facts so I gave them to you... David is an Amateur. I ma happy that you are such the expert... If I look a most of your response you are a TROLL...

[/quote]
Certainly the name calling and harrassment will inspire him to respond. Good idea.[/quote]

You asked for facts so I gave them to you... David is an Amateur. I ma happy that you are such the expert... If I look a most of your response you are a TROLL...

As an honest-broker and third-party, I'm just interested...no...intrigued at how, whenever things go wrong, all these people claiming to be "network admins" and "system engineers" suddenly rush out of the woodwork to give us their two-cent diatribes on what went wrong and how "I would NEVER let this sort of thing happen to me!" It gets even more interesting when they call the people who are running this "beginners", even though it's still the same incarnation since 1996's first alpha test of the program from conception to program. The kicker, though, is that it's the same thing like Microsoft and all of the computer technology in that nothing works 100% of the time. Does that mean the people at Microsoft are all hack newbies? Does that mean all the designers of missiles, planes, cars, trains, and everything else that fails every so often are insufficient?

Just an observation. I'm sure those people up there, by your logic, better spend their time working out the kinks of the program and not spend their time talking on the forums to answer every single question from the masses,(ala you).

---

Certainly the name calling and harrassment will inspire him to respond. Good idea.[/quote]

You asked for facts so I gave them to you... David is an Amateur. I ma happy that you are such the expert... If I look a most of your response you are a TROLL...

As an honest-broker and third-party, I'm just interested...no...intrigued at how, whenever things go wrong, all these people claiming to be "network admins" and "system engineers" suddenly rush out of the woodwork to give us their two-cent diatribes on what went wrong and how "I would NEVER let this sort of thing happen to me!" It gets even more interesting when they call the people who are running this "beginners", even though it's still the same incarnation since 1996's first alpha test of the program from conception to program. The kicker, though, is that it's the same thing like Microsoft and all of the computer technology in that nothing works 100% of the time. Does that mean the people at Microsoft are all hack newbies? Does that mean all the designers of missiles, planes, cars, trains, and everything else that fails every so often are insufficient?

Just an observation. I'm sure those people up there, by your logic, better spend their time working out the kinks of the program and not spend their time talking on the forums to answer every single question from the masses,(ala you).
[/quote]
Obviously you missed on key point

The single point of failure other than the already broken network issue is the Scheduler. how many times have you seen the Scheduler is not responding in RED

Ned asked for points so I expanded.. he then came back, I added facts.. These are things that are littered through the forums...

Why David does not have time to read and then have the limited staff attempt to correct issues (even if it takes months) is beyond me... There are many IT Professional peple that have been members for Ages and at times get very tired of the Computer mistakes that appear on a regular basis. Teh question becomes why shuld they keep happening? From and IT standpoint there are only two reasons.

1. we have no money
2. we do not know how to fix the problem. If we reboot the computer it will run along fine until it happens again.

IF we we do not say anything the users that donate their time will be very upset (Some will quit) but we will get new ones to replace them.. NO Big Deal.

IF we keep the ruse going long enough, others will volunteer help and keeps a few for a while longer.

Paul Buck has did more for what Seti Should have Documented... That is only part of a project "documentation." As I said Thank You Paul Buck!

So if that is being a TROLL then I guess that I am now a TROLL.

I now have to set a a schedule to come out from under my bridge and extract a toll on the users of Seti!

I appoligize that you were drawn into this mess..

Al

I think bad things tend to happen during the weekends because less people are paying attention to the servers and therefore not fixing problems right away (though I did spend two hours last night screwing around with the web server which was all hung up).

But in this case, I think there is a bunch of minor problems caused by the recent web domain switch which in turn caused more demand on the whole set of BOINC servers. Basically this is yet another BOINC stress test. One odd thing is that it is too difficult to get a news item posted on the front page (since the machine hosting the pages is having NFS issues I don't want to deal with right now).

And it is the weekend, and I have plans, so.. see you later.
- Matt

Hi Matt .. thanks very much for coming in Saturday June 4 your weekend off and working hard to get the system up and running again and thanks for your post and information . have a good weekend what's left of it I hope you have a Gig lined up and get to play some Good music . like the one you had on the USS Midway. The one that's now the "Midway Museum." Big party for an international conference of about 2-3000 lawyers.
Regards ... byron ... beta tester ... of Boinc ... since ... July 12 , 2003 ... until public launch of .... SETI@hone 2 / Boinc ... June 22 , 2004

But then what do you expect from "Amateurs"

Al

Its all well and good calling people names but have you considered that the science is the main consideration rather than spending every donated penny on triply redundant 100% uptime systems for the distributed computing aspect just to keep people crunching at the cost of the actual science being done with the results that are contributed?

I am sure they will be more than happy for you to donate a redundant backup network, all the servers, operating costs and experience for it, if you are unable to do so, I would suggest that you help them find someone who can rather than dissing the hard working team because they havent got the ability to produce duplicate hardware, network and operating costs out of thin air.

In response to your comment about office networks run by secretaries, any sysadmin worth the title would know there is a major difference between a massive distributed computing network and a windows NT file server sat on a spare box in the corner. With unlimited resources anyone can manage a network effectively, however when those resources are not available then even the best sysadmins in the world will have some difficulties and occasional outages.

---

Part of Teamseti
For SetiBoinc status graphs visit Teamseti status graphs

[font='fixedsys,courier']I hope Pappa is wearing asbestos BVD's...
- Keeper of the Flames[/font]

---