There is a lot of fuming and foaming going on in this forum about the down time, and I have to assume that the SETI folks are working hard behind the scenes to set things right.

But I would like to see more frequent communication of status and expectations than has been the case. For example, the last status post -- which was made about 24 hours ago -- says that things are now supposedly working, when for most of those same 24 hours it is quite clear that things are not working.

Doing a better communication job would help a lot.

> There is a lot of fuming and foaming going on in this forum about the down
> time, and I have to assume that the SETI folks are working hard behind the
> scenes to set things right.
>
> But I would like to see more frequent communication of status and expectations
> than has been the case. For example, the last status post -- which was made
> about 24 hours ago -- says that things are now supposedly working, when for
> most of those same 24 hours it is quite clear that things are not working.


That's the UCB norm since they made Seti/Boinc live! The system has not been up for a 72 hr period since the launch on 6/22. Infact, that was the norm in the 7 months I was with the beta. I think we need to form a team called "mushrooms"

Not an excuse, by any means, but the "New Coke Crew" just doesn't get it!

> The system has not been up for a 72 hr period since the launch on 6/22.

I'll call you on this, your just spewing crap to spew crap.


----- Rom
BOINC Development Team, U.C. Berkeley

> There is a lot of fuming and foaming going on in this forum about the down
> time, and I have to assume that the SETI folks are working hard behind the
> scenes to set things right.
>
> But I would like to see more frequent communication of status and expectations
> than has been the case. For example, the last status post -- which was made
> about 24 hours ago -- says that things are now supposedly working, when for
> most of those same 24 hours it is quite clear that things are not working.

Well when I posted that message I thought I had things well in hand, but it takes awhile for the system to play catchup.

What I thought would clear up the backlog turned out to be a net gain of zero, as bringing up another transitioner added additional load to the database such that it slowed down the other transitioners. At 7am this morning I removed the additional transitioner and reverted us back to 4 transitioners.

Any futher progress was hampered today because the site was down or slowed down due to backup processes going on in the background.

I continued to investigate ways of bringing down the database server load through the use of batching requests, while another group of people are preparing to move the database to another storage medium that should increase our overall system throughput by a factor of 5 or 6.

During the move the system will more than likely be down for a few hours. I don't know when the Ops side will be ready for the move though.

Meanwhile we have been providing support for the various other project on server configurations and system configuration issues which seem to be causing different problems for them.

I've also been building 4.01 for the Alpha folks and debugged an issue reguarding benchmarks and heartbeats to the project applications which caused the project applications to terminate themselves. I've checked in that change and now and rebuilding all the core clients.

----- Rom
BOINC Development Team, U.C. Berkeley

@ AZ Woody.

Just like in beta, you are doing a great job of establishing yourself as a FLAMER. The only difference between you and JQ is that JQ didn't pretend to be a know-it-all.


Boinc ALPHA/BETA Tester

---

Account frozen...

Thanks very much for the update, Rom. I thought you guys must be working feverishly, and so you have been. Hang in there ...

OK, here it is in a nut shell. The equipment currently allocated for the Boinc release is just a small portion of the hardware available. The bulk of the hardware is still tied up running S@H1. The original release of the Boinc client to the public was supposed to be a phased in operation. However, some eager beaver let the cat out of the bag by releasing the startup site URL. Some decisions had to be made and it was decided to finally give it a go while the tidal wave was approaching. Unfortunately the available hardware was not yet optimized to run so many hosts, and with more and more hosts trying to come on line, it has become a constant game of catch up. The obvious solution would be to shut down S@H1 and divert all that hardware to Boinc. This however would leave a mass of hosts at the old site in the lurch.

To make a long story short, it is going to be a rough road for little while, until compromises can be worked in and the full transition made. Patience is going to be needed. The systems will constantly be needing adjustment which means that it will be going down and coming back up again rather frequently for some time. Eventually, with all the available hardware on line at Boinc and optimized, the situation will radically improve. It will just take some time.

P.S. Don't ask anyone how long it will take, we do not have the ability to see into the future.

P.P.S. You can bet your last dollar that some mistakes will be made, but until you can establish your godhead and prove that you can walk on water, keep the critical comments civilized and to a minium unless you can provide proof of your own perfection.




Boinc ALPHA/BETA Tester

---

Account frozen...

Amen to that Dogbytes, a very nice, accurate, and to the point post.

(-: FireFox :-)

You guys keep up the excellent work and things will come together sOONtm.



> > There is a lot of fuming and foaming going on in this forum about the
> down
I will agree to disagree that all 15 of the foaming and fuming forumites are real representative of the the total population of BOINC ;0).....Pascal the ornery oldman


> > time, and I have to assume that the SETI folks are working hard behind
> the
> > scenes to set things right.
> >
> > But I would like to see more frequent communication of status and
> expectations
> > than has been the case. For example, the last status post -- which was
> made
> > about 24 hours ago -- says that things are now supposedly working, when
> for
> > most of those same 24 hours it is quite clear that things are not
> working.
>
>
> That's the UCB norm since they made Seti/Boinc live! The system has not been
> up for a 72 hr period since the launch on 6/22. Infact, that was the norm in
> the 7 months I was with the beta. I think we need to form a team called
> "mushrooms"
>
> Not an excuse, by any means, but the "New Coke Crew" just doesn't get it!
>
>

Please Woody you are making me laugh so hard I spit Pepsi all over my keybd and it was thru my nose, so I beg of you no more funny stuff, my heart can not take it hehehehehehelolololololol

Owner's manual

Website Owner's Manual

He is not a flamer, he his a wantabe TROLL lolol

Bah!!! I am in it for the money.

Message from the Head ;o) of the
Horse Head Nebula Branch of Seti@Home,

Check is in the mail

> Well when I posted that message I thought I had things well in hand, but it
> takes awhile for the system to play catchup.
>

Thanks for keeping us informed matey!

Owner's manual

Website Owner's Manual
Bah!!! I am in it for the money.

Message from the Head ;o) of the
Horse Head Nebula Branch of Seti@Home,

Check is in the mail