Post your S5R1 completion estimates here

Bernhard Frey
Bernhard Frey
Joined: 1 Apr 05
Posts: 1
Credit: 230812
RAC: 0

RE: Just taken a look at

Message 45503 in response to message 45502

Quote:
Just taken a look at the server staus page and it says that 1.4 million unit validated in the last week and only 4.2 million units remaining. Going from those figures it would suggest we only have another 3 weeks of S5R1 to go. Thats about the 10th of Jan.

A little mistake in your calculation. There are 1.4 million results (not units), that have been validated last week. Since a workuint consits in average of 2.2 results (2 initial replication + a guess of 10% resends for resultes not comming back in time) it would be around 6,6 weeks to go. This is still a bit less than the 63.6 days to go estimated in the server status page.

My bet is that we'll fall on 14th of February below 10.000 units left (and will have another 6 weeks to go until the last S5R1 WU was processed)

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6534
Credit: 284700169
RAC: 113726

RE: RE: When S5R1 is done

Message 45504 in response to message 45501

Quote:
Quote:
When S5R1 is done will we have a dry spell until the next big run or will more work be available immediately?

Bernd already said that S5R2 is being written and tested on their end. So we should be good to go, I would think.


Listening to the AEI presentations, the comment was made ( M.Papa ? ) that even during the multi-year hiatus of collection while LIGO is upgraded to advanced, then number crunching at E@H will continue using deeper searching of even wider data sets than we are doing now.

Cheers, Mike.

( edit ) It was also emphasised several times how crucial the E@H effort is to the LIGO project - E@H participants contribute ~80 TFlops of computing power which enable any sense to be ever made of the raw data output of the IFO's.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

RandyC
RandyC
Joined: 18 Jan 05
Posts: 6003
Credit: 111139797
RAC: 0

RE: RE: Just taken a look

Message 45505 in response to message 45503

Quote:
Quote:
Just taken a look at the server staus page and it says that 1.4 million unit validated in the last week and only 4.2 million units remaining. Going from those figures it would suggest we only have another 3 weeks of S5R1 to go. Thats about the 10th of Jan.

A little mistake in your calculation. There are 1.4 million results (not units), that have been validated last week. Since a workuint consits in average of 2.2 results (2 initial replication + a guess of 10% resends for resultes not comming back in time) it would be around 6,6 weeks to go. This is still a bit less than the 63.6 days to go estimated in the server status page.

My bet is that we'll fall on 14th of February below 10.000 units left (and will have another 6 weeks to go until the last S5R1 WU was processed)

There's a little mistake in your calculation as well. At 6.6 weeks (46.2 days) remaining from date of your post (19 Dec), that gives an end date of about 03 Feb. 2007.

However, we've been returning an average of 132500 results per day over the last 4 days. At that rate (which is not guaranteed to continue by any means), we should finish in about 31 days from today (20 Dec 2006)... i.e. around 20 Jan. 2007.

Seti Classic Final Total: 11446 WU.

RandyC
RandyC
Joined: 18 Jan 05
Posts: 6003
Credit: 111139797
RAC: 0

Dispite the holidays and the

Dispite the holidays and the recent downtime, we still seem to be on pace to complete (i.e. send out the last WU) sometime around 18-19 January 2007.

Status as of 12:25 PM UTC on Sunday, 31 December 2006 - 2,551,743 WU remaining

Seti Classic Final Total: 11446 WU.

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6534
Credit: 284700169
RAC: 113726

RE: Dispite the holidays

Message 45507 in response to message 45506

Quote:

Dispite the holidays and the recent downtime, we still seem to be on pace to complete (i.e. send out the last WU) sometime around 18-19 January 2007.

Status as of 12:25 PM UTC on Sunday, 31 December 2006 - 2,551,743 WU remaining


It's exciting, isn't it?
As recent reports have indicated, the total processing power of E@H and the processing output ( science wise ) is rising.
My gut feeling/optimism sense quite a momentum to the project, as despite some troubling setbacks, the progress is still good.
Just you wait and see what happens when the first ever signal detection is given hard confirmation!! I can't wait.

Congratulations for the excellent work by the project staff, and a big thankyou to all E@H participants!!

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Svenie25
Svenie25
Joined: 21 Mar 05
Posts: 139
Credit: 2436862
RAC: 0

I think, it could be a litlle

I think, it could be a litlle bit faster, when LHC@ doesn´t sent so much work in the last few days... ;)

archae86
archae86
Joined: 6 Dec 05
Posts: 3145
Credit: 7023044931
RAC: 1831966

RE: As recent reports have

Message 45509 in response to message 45507

Quote:


As recent reports have indicated, the total processing power of E@H and the processing output ( science wise ) is rising.

Cheers, Mike.


I've been wondering why the percent per day graph is rising so steeply, clearly out of sync with any measure of available computing power.

I finally noticed a minute ago that the status page appears clearly to specify this is percent of _units_ complete--without short/long weighting.

As many have observed, units currently being handed out are very short-rich compared to even a few weeks ago.

If this stays true, we'll scorch along to the end much faster than I thought.

RandyC
RandyC
Joined: 18 Jan 05
Posts: 6003
Credit: 111139797
RAC: 0

RE: As many have observed,

Message 45510 in response to message 45509

Quote:


As many have observed, units currently being handed out are very short-rich compared to even a few weeks ago.

If this stays true, we'll scorch along to the end much faster than I thought.

You have a good point. It appears that the number of long vs short WUs is getting smaller rapidly (i.e. we're running out of long WUs). Bruce posted on this earlier here

So the fact that we're completing more WUs per day is not necessarily an indication of a massive increase in computing power, but reflects more on the length of WUs now being distributed/returned.

Seti Classic Final Total: 11446 WU.

Arion
Arion
Joined: 20 Mar 05
Posts: 147
Credit: 1626747
RAC: 0

I've been gettign tons of

I've been gettign tons of those darned small wu's too. Normally I wouldn't care, but with the server going down and I have 3 systems slamming away trying to dump 60 to 80 packets its really messing up my internet connection. Haven't figured out how to stop all those completed units from trying to go into the void for credit.

On another note.... what happens when we run out? I havent' understood what you all have been talking about a new project or ????

Arion

Alinator
Alinator
Joined: 8 May 05
Posts: 927
Credit: 9352143
RAC: 0

About all you can really do

About all you can really do with the current BOINC clients is kill network connectivity or suspend the project until the trouble goes away. People have suggested the allowing the client to selectively disable the connection for individual projects would be a nice feature to have.

Another workround would be to create an advanced rule for your firewall to block a project IP block or domain when they go down. This would at least stop the futile traffic from making it to the wire, but may still have a local impact from the User's POV. Another idea if you have your own DNS server running would be to temporarily route the project to a "blackhole" IP.

Alinator

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.