It seem that Seti@home is functioning properly again. I don't know if this applies to everybody but i will share what i just done to get your computer to start crunching Seti@home work units again.

This has worked for me and i now have a full load of work units crunching again!

If BOINC manager is not uploading or downloading any work units, you have to delete BOINC off your computer and download it again. This will get you going again.

****Simple instructions for beginners*****
This is with windows xp because that is what i have running.

1. Click Windows Start Menu-> my computer-> C: ->program files -> Right click BOINC and delete it!

2. Visit SETI@home main page and download the latest version of BOINC. Here is a link
http://boinc.berkeley.edu/download.php

3. Run the BOINC software and when prompted enter the URL or Internet address; http://setiathome.berkeley.edu

4. When asked if you are a new or existing user, select existing user and put in your e-mail address and password.

5. BOINC manager will then connect to SETI@home and get you new work units.


I hope this helps some people as the SETI@home front page does not have much help on it.

John in Ireland.

---

John's method will have important side-effects for some users:

It will delete all work in progress, including for other projects you are connected to, apart from SETI
It will delete any local statistics and logs stored on your computer (it won't affect stats on the SETI website, or any third party sites)
It will delete any optimised or other special applications on your system.

So if you have the simplest of installations, and don't crunch for any project except SETI, John's method is fine.

For everybody else - stop and take a deep breath first, and think whether you really want to do that. Other methods are available.

Which methods?

Kelwin since 99

Which methods?

Kelwin since 99

Try this one, tested and recommended by Simon the Great Chicken yesterday.

Which methods?

Kelwin since 99

Try this one, tested and recommended by Simon the Great Chicken yesterday.

I have checked the chickens, they are not charing with Old Chomputer Chusers..
K.

It seem that Seti@home is functioning properly again. I don't know if this applies to everybody but i will share what i just done to get your computer to start crunching Seti@home work units again.

This has worked for me and i now have a full load of work units crunching again!

If BOINC manager is not uploading or downloading any work units, you have to delete BOINC off your computer and download it again. This will get you going again.

****Simple instructions for beginners*****
This is with windows xp because that is what i have running.

1. Click Windows Start Menu-> my computer-> C: ->program files -> Right click BOINC and delete it!

2. Visit SETI@home main page and download the latest version of BOINC. Here is a link
http://boinc.berkeley.edu/download.php

3. Run the BOINC software and when prompted enter the URL or Internet address; http://setiathome.berkeley.edu

4. When asked if you are a new or existing user, select existing user and put in your e-mail address and password.

5. BOINC manager will then connect to SETI@home and get you new work units.


I hope this helps some people as the SETI@home front page does not have much help on it.

John in Ireland.

As an update to my last post, i have now finished my first 2 work units and they have been returned to SETI and are now just pending their credit!

So this method will work and get you crunching again!

John in Ireland.

---

Can someone please tell me in simple english why SETI is not downloading any work from the project?

The Berkeley (project) servers are broken.

As I don't think the servers are broken any longer.

The servers themselves are starting to settle down; the problem is in the new server code that was deployed recently.

I think “broken” covers that as well—in the simplest possible ‘non-geek’ terms, considering “the servers” to comprise hardware and software components both.

---

Hey what about the old machine 'Thumper', the one which conked off and started all this. Only the motherboard is faulty in that machine right ?

If so can't it fixed, i mean now we have time on our hands now with the replacement online but if the old machine can be fixed we'll have more server.

Chip level repairs are possible but if in this case they are not, you can place an AD in the newpapers or somthing or even contact SUN for a new motherboard.

If we can get that machine working again, it would be great and its definitely worth a shot.

What do you think ?

The old "Thumper" was basically one-of-a-kind hardware -- an early engineering unit that was good, but not something where Sun would have spares available for purchase.

So, in a way, the death of Thumper is a very good thing: the replacement is 100% production components, and motherboards can be bought if needed.

---

If I recall correctly, the failure on Thumper was with the drive controller, couldn't it still be of some use as a splitter or server running discless?

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

If I recall correctly, the failure on Thumper was with the drive controller, couldn't it still be of some use as a splitter or server running discless?

I don't know.

What I remember reading is that Thumper was built as a stack of boards, with the drives plugging directly into a PC Board -- no cables at all (and the new ones are built the same way, but the dimensions are slightly different).

Probably the best way to use the old chassis is to "salvage" the high-value parts and use them on commodity boards.

---

If I recall correctly, the failure on Thumper was with the drive controller, couldn't it still be of some use as a splitter or server running discless?

I don't know.

What I remember reading is that Thumper was built as a stack of boards, with the drives plugging directly into a PC Board -- no cables at all (and the new ones are built the same way, but the dimensions are slightly different).

Probably the best way to use the old chassis is to "salvage" the high-value parts and use them on commodity boards.

But if the only failure was the ability to see the drives, if it will boot and the comms were not compromised, it would still have a lot of processing/server power, no?

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

If I recall correctly, the failure on Thumper was with the drive controller, couldn't it still be of some use as a splitter or server running discless?

But if the only failure was the ability to see the drives, if it will boot and the comms were not compromised, it would still have a lot of processing/server power, no?

msattler,
take a look at what Thumper looks like. Actually the new hardware has two low profile expansion slots, but you can't be sure if the preproduction sample had these, too. Better rip out the usable parts as spares.

---

_\|/_
U r s

As of this post -
This thread Viewed 10,000 times !!

We are needing some sort of new update I think :)

Seems like it!!

---

It is no good to try to stop knowledge from going forward. Ignorance is never better than knowledge. --- Enrico Fermi ---

Hey what about the old machine 'Thumper', the one which conked off and started all this. Only the motherboard is faulty in that machine right ?

If so can't it fixed, i mean now we have time on our hands now with the replacement online but if the old machine can be fixed we'll have more server.

Chip level repairs are possible but if in this case they are not, you can place an AD in the newpapers or somthing or even contact SUN for a new motherboard.

If we can get that machine working again, it would be great and its definitely worth a shot.

What do you think ?

Thumper has already been replaced, and up and running. You can read that on the front page news as well as in some of the other Tech News items here. Its a lot more deeper than I understand, and a lot of it has been explained by Matt pretty much every day he works. Check out some of the other threads, and you'll see. Hope that helps.

Jeremy

We all know thumper is replaced
but the only thing wrong with it is the "sata controlers"
we got perfectly good working
Ram
Cpu's
mainboard (exept sata)
Harddrive's (from new thumper)
if we can get an alternative way to get at least 1 hard drive going it can make an mean spliter,upload/download server,sceduler
if we cand get old chasis working
get an barebone chassis and put old thumper mem and cpus + HDD's in there (nice replica master science DB)
Just my 2c

Conrad

Conrad,

If you look at the person who I was responding to, thats why I said what I said. I am well aware of the new things needed, its been said by Eric and Matt. He simpy was asking about a replacement for thumper, and I simply stated it has already been replaced. Please read quoted items before responding so people don't get confused. Thank you.

---

Probably the best way to use the old chassis is to "salvage" the high-value parts and use them on commodity boards.

But if the only failure was the ability to see the drives, if it will boot and the comms were not compromised, it would still have a lot of processing/server power, no?

It depends alot on what you think of as "server power."

That usually isn't processor power, it's usually memory and DISK.

IMO, losing the ability to see disks is a pretty big loss.

---

Keep getting error messages like this group:

5/19/2007 3:47:59 AM|SETI@home|Scheduler request failed: HTTP internal server error
[snip]

I have shutdown and restarted Boinc and still get these messages.

If you are running KWSN's optimized apps, then read in this thread: http://setiathome.berkeley.edu/forum_thread.php?id=39531. If you have already followed the instructions around app_info.xml and have got work, then you might consider using the command "No new work".

---

Kind regards
Kirsten

Third (my local solution) suspend SETI each Tuesday morning and resume SETI Wednesday morning.

I don't watch my PC's closely enough to know when there are minor SAH outages. If this is consistent enough and enough people agree in the concept of a Tuesday am-pm shutdown then I have an idea.
First, I'm sure that making an architectural change in BOINC or SAH is out of the question to deal with a local (though often) problem.

What about *someone* creating an application that would run in the background (Winodws service, taskbar app, command-line ?) that would simply shut down BOINC at the configured time and then start it back up at the configured time?
I don't think the app would need to be supported by the SAH team nor be a burden on them.

Perhaps there is a generic tool like this already around - I haven't looked. I'm trying to think of something that would be very simple to install, run and manage so it would be simple to deploy. Yes, you could do it with shell scripts, batch files/resource kit tools too.

Good idea? Bad idea?

Joe

(I hope I got the bracket quote bracket thingies right)

So what that would entail is a sched (chron) job to send the SUSPEND and RESTORE messages to SAH for the Tuesday shutdown. Unfortunately, most of MY programming experience has been in ALGOL on Burroughs mainframes. This would require more knowledge of Windows (or Linux / UNIX) and the BOINC protocol than I have.

Well for Windows, batch files containing the "net start boinc" and "net stop boinc" commands, run in scheduled tassks would be easiest i reckon, provided the user installed as a service.

I am sure it could work for running not as a service, however terminating boinc would be more difficult, but could use boincmd.exe functions to control processing / workunit fetching etc.. instead

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

It depends alot on what you think of as "server power."

That usually isn't processor power, it's usually memory and DISK.

IMO, losing the ability to see disks is a pretty big loss.

Yep, looking at the link posted above, it doesn't appear possible to drop in a SATA card to give it at least one HDD- there just isn't any room (other than some serious cut & hack of the drive storage area) to mount any HDDs .
If it's able to boot from the network then it'd be suitable for anything that doesn't require a lot of disc throughput using the ethernet port(s) for it's storage needs.

---

Grant
Darwin NT

Several users here in the Seti@Netherlands community report that removing optimized seti clients completely (inc. app_info.xml) and restarting from 0 with fresh downloaded original seticlient restores connectivity with Berkeley to nearly normal level.