Author Topic: Seti is down again (Read 608628 times)

JohnDK · « **Reply #570 on:** 27 Oct 2010, 06:16:43 pm »

Quote from: Raistmer on 27 Oct 2010, 05:54:29 pm

Forums online now

Was that 5 mins uptime? All seems down now...

perryjay · « **Reply #571 on:** 27 Oct 2010, 06:22:34 pm »

Dang it, I must have blinked!! I missed it.

JohnDK · « **Reply #572 on:** 27 Oct 2010, 07:56:46 pm »

Quote from: perryjay on 27 Oct 2010, 06:22:34 pm

Dang it, I must have blinked!! I missed it.

Quick it's up again

perryjay · « **Reply #573 on:** 27 Oct 2010, 09:37:36 pm »

Got it this time. Jeff Cobb posted over in the Tech news Forum

Quote

Just a quick note. Obviously, jocelyn is up. Mork is recovering.

The purchase orders for both oscar and the new mork went out late today or will go out early tomorrow. It takes a while for these things to work their way through the purchasing pipeline.

We decided to go with HP for these machines. They gave us a very good deal. We are getting two identical (oscar class) machines. I'll post the specs in another note. We hope to have them on hand in about 2 weeks.

At this point, we are discussing what we will do between now and when the new servers are on line.

Miep · « **Reply #574 on:** 28 Oct 2010, 03:06:39 am »

I hope you can all appreciate the irony that people are actually asking not to get work

Gizbar · « **Reply #575 on:** 28 Oct 2010, 05:03:53 am »

Quote from: Cosmic_Ocean on 27 Oct 2010, 05:04:18 pm

Well it's not a ridiculously old board.. only two generations old. It's a K8, but it's a socket 754 board. I got it almost four years ago just before newegg quit selling s754 boards. Needed an office workstation, so $300 in parts built an entire machine.

Newer boards still have caps, but they're the small compact ones, and the high quality Japanese caps that basically don't ever go bad, so "bad cap" replacement is likely nowhere near common anymore.

Yes, the newer 'solid state' caps that they're generally called, seem to be a lot more reliable. Not that they are 'solid state'. Never had one of those 754's, went from socket 462 to 939, then on to AM2/AM3.

Gizbar · « **Reply #576 on:** 28 Oct 2010, 05:12:07 am »

Quote from: Miep on 28 Oct 2010, 03:06:39 am

I hope you can all appreciate the irony that people are actually asking not to get work

Well I still want work! It's getting cold in here!

Raistmer · « **Reply #577 on:** 28 Oct 2010, 05:25:07 am »

Einstein@home (CPU), MilkyWay (ATI) and Collatz (NV) are keeping my room warm

[Actually I have lot of ATI work for SETI still, just can't report it. Hope this time "hour of MW" will not come

]

Gizbar · « **Reply #578 on:** 28 Oct 2010, 03:17:03 pm »

Quote from: Raistmer on 28 Oct 2010, 05:25:07 am

Einstein@home (CPU), MilkyWay (ATI) and Collatz (NV) are keeping my room warm
[Actually I have lot of ATI work for SETI still, just can't report it. Hope this time "hour of MW" will not come ]

I just can't bring myself to sign up for another project, seeing as Seti was all I ever wanted to do... Don't really want to sign up to a project that will get abandoned as soon as Seti sorts itself out!

Just got 2 AP left to report, everything else finally went through.

Down to playing games now, but that just doesn't make the computer as warm as 6 seti wu's!

Not to mention one rig shut down for days now! I might even have to put the heating on soon!

Richard Haselgrove · « **Reply #579 on:** 28 Oct 2010, 03:29:22 pm »

No shame in attaching to a project for a short time, and detaching again afterwards - just please allow any work in progress to flush through (or at least report the work as aborted) before you leave. Probably best not to choose CPDN, then

In the meantime,

Quote from: David Anderson

I added a new config option, <dont_send_jobs>.
When set, the scheduler doesn't send jobs,
and the feeder doesn't enumerate them.

Set this during periods when you want to let your server
catch up handling completed jobs, and not issue new ones.

I wonder which project might be feeling the need to let its servers "catch up handling completed jobs", LOL

Claggy · « **Reply #580 on:** 28 Oct 2010, 03:33:49 pm »

Quote from: Richard Haselgrove on 28 Oct 2010, 03:29:22 pm

No shame in attaching to a project for a short time, and detaching again afterwards - just please allow any work in progress to flush through (or at least report the work as aborted) before you leave. Probably best not to choose CPDN, then

In the meantime,

Quote from: David Anderson
I added a new config option, <dont_send_jobs>.
When set, the scheduler doesn't send jobs,
and the feeder doesn't enumerate them.

Set this during periods when you want to let your server
catch up handling completed jobs, and not issue new ones.

I wonder which project might be feeling the need to let its servers "catch up handling completed jobs", LOL

I was wondering that too, i hope resend lost tasks still works with the option enabled,

Claggy

Edit: I've been busy doing CPDN, Einstein and Collatz on my E8500/9800GTX+/HD5770 machine, Astropulse and CPDN on my T8100 Laptop, and Einstein on my Atom Netbook,
i always tend to keep a couple of Suspended CPDN Wu's on the E8500 and T8100 for extended server outages, with Resourse share set to 15% of Seti's.

Claggy · « **Reply #581 on:** 28 Oct 2010, 06:17:59 pm »

Jeff has posted this:

Quote

We've decided to keep the project down until the new servers are up and running and the databases migrated to them.

The forums will stay up.

The back end and the upload server will stay up until we clear the outstanding results.

The time line we are looking at is about one month - two weeks for the servers to arrive and another two to get them going. We'll see as time goes on whether or not that's too aggressive.

The down time will be used for preparing the databases for migration. For example, on the science side, we can finally finish a big merge of the spike table and drop the old spike table. This will make the database smaller and easier to migrate.

We will also use the time for science processing and analysis.

More later...

Claggy

Josef W. Segur · « **Reply #582 on:** 28 Oct 2010, 11:04:57 pm »

Quote from: David Anderson

I added a new config option, <dont_send_jobs>.
When set, the scheduler doesn't send jobs,
and the feeder doesn't enumerate them.

Set this during periods when you want to let your server
catch up handling completed jobs, and not issue new ones.

Quote from: Claggy on 28 Oct 2010, 03:33:49 pm

... i hope resend lost tasks still works with the option enabled,

Claggy

Sorry, no such luck.
Joe

arkayn · « **Reply #583 on:** 29 Oct 2010, 01:30:09 am »

I decided to let my 2 Quads run some Aqua for a month or so, the iMac will just get by with Collatz.

msattler · « **Reply #584 on:** 29 Oct 2010, 03:49:10 am »

I just cut the Frozen One over to MW.......makes me a little sick having to do so, but otherwise I would have to start the furnace to keep the kitties warm.