Hello,

Does the message regarding weekly outages mean that new work units are not available three days out of the week? I have set my preferences to maintain five days worth of work, but this never happens. I have not had good luck getting backup projects to 'play nice' with SETI - they seem to want to take over. I'm a little stuck between trying to keep BOINC busy and getting the most SETI possible.

---

Hello,

Does the message regarding weekly outages mean that new work units are not available three days out of the week? I have set my preferences to maintain five days worth of work, but this never happens. I have not had good luck getting backup projects to 'play nice' with SETI - they seem to want to take over. I'm a little stuck between trying to keep BOINC busy and getting the most SETI possible.

There are several threads in the Number Crunching forum that describe what happens with the weekly shutdowns. The root of it is that during the outages, you can not download new work, or upload your finished work. There are so many posts that describe portions of it that I'm not sure where to point you except for the Number Crunching Forum itself.

http://setiathome.berkeley.edu/forum_forum.php?id=10

Welcome to the SETI boards!

Steve

---

Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website

Okay, thanks for that Steve. I guess what it's down to is getting enough units to last the three days. Can anyone suggest how much disk space I need per work unit. I can then make sure my preferences are set with this in mind.

---

Mine only uses 8.2 MB

---

Pluto will always be a planet to me.

Seti Ambassador
Not to late to order an Anni Shirt

I am using about 120 meg but I have another project I crunch for that has far bigger work units.

---

Thanks Uli and Dena. My settings should be more than ample as far as disk space, in that case. Are there other settings to which I should pay special attention with respect to getting a reserve of work units, does anyone know?

---

The number of days of work you wish to keep on hand.

---

Make sure you have a four to five day cache to make it through the outage and I recommend you suspend network activity during that time. I generally don't unsuspend mine until Friday evening. I am in the same timezone as Berkeley.

---

Pluto will always be a planet to me.

Seti Ambassador
Not to late to order an Anni Shirt

Just select another BOINC project until the the problems are solved.

---

http://boinc.mundayweb.com/teamStats.php?userID=14824

Just select another BOINC project until the the problems are solved.

You should be aware that adding other projects may introduce new issues. Some projects have a range of completion times, and a single WU download can introduce a large demand from one project, slowing or preventing downloads from other projects. This will all sort itself out eventually, but it can take several weeks.

For example, a few weeks ago WCG requested and received more WU than usual on one of my machines. I have no idea why. As a result, WCG went into high priority mode on both CPUs for a few days, and I missed getting enough S@H downloads for the first 3 day outage. All my CPUs were busy through the outage, just not with the expected split between projects.

---

Thanks for your thoughts, Bill and Yank. I had already run into problems with Climate Prediction, which has massive work units with ungenerous deadlines. It has been out of the picture for a few weeks now.

I am reluctant to add another project until I have SETI running the way I want. I set my preferences to download 5 days worth of work 2 - 3 weeks ago, but I have yet to see more than three work units on my PC at any given time, and even then one of them is usually waiting to report.

Still searching for the missing ingredient...

---

Still searching for the missing ingredient...

Perhaps that's clearing the local preferences?

Gruß,
Gundolf

---

Computer sind nicht alles im Leben. (Kleiner Scherz)

SETI@home classic workunits 3,758
SETI@home classic CPU time 66,520 hours

Indeed, no project is 100% complication free and 100% reliable all the time.

Amongst my set of projects, I've found that Einstein and POEM are among the best, with World Grid, Spinhenge and Climate clost behind. On the GPU side, as they are honestly small and truly resource constrained (none of those projects have resources (or user bases) remotely approaching the big boys like SETI and Einstein), there tend to be a few more bumps in the road. But they generally run reasonably reliably and tend to be fairly good at communication.

Mixing and matching projects can be a bit of a chore -- I suspect part of that has to do with the projects themselves, and part has to do with the BOINC client design -- but it can be managed reasonably well.

The thing is, for the moment, none of the other projects I work with have a set three day plus outage in their weekly operation. I think the good news here is that for a long time SETI has been the sugar attracting a large new user base to shared computing -- so with the current SETI 40% outage rate, those new users (and for that matter long time SETI only users) may find some amount of shared computing solace attaching to one or more of the large number of worthy (and currently more highly functioning) alternative projects out there.

You should be aware that adding other projects may introduce new issues. Some projects have a range of completion times, and a single WU download can introduce a large demand from one project, slowing or preventing downloads from other projects. This will all sort itself out eventually, but it can take several weeks.

For example, a few weeks ago WCG requested and received more WU than usual on one of my machines. I have no idea why. As a result, WCG went into high priority mode on both CPUs for a few days, and I missed getting enough S@H downloads for the first 3 day outage. All my CPUs were busy through the outage, just not with the expected split between projects.

---

Thanks for your thoughts, Bill and Yank. I had already run into problems with Climate Prediction, which has massive work units with ungenerous deadlines. It has been out of the picture for a few weeks now.

I am reluctant to add another project until I have SETI running the way I want. I set my preferences to download 5 days worth of work 2 - 3 weeks ago, but I have yet to see more than three work units on my PC at any given time, and even then one of them is usually waiting to report.

Still searching for the missing ingredient...

Weird thing is my P4 is doing a climate and I have had no problems with it uploaded my last trickle around 19:00 BST which is UTC + 1

---

i also have a p4(d) Chugging away doing two CP models at once doing the occasional trickup - last one 15 minutes ago (20:15 bst)

one has 5 days to go and the other 15 days to go (cpu time) :(

---

Tim

Maybe I did not explore my options on Climate Prediction enough, because the work units I got took months on my 2.8 GHz machine which is idle 20 hours a day, or more. The deadlines were so tight that BOINC was always giving them priority, so I finally just took CP out of the picture.

At any rate, I am going to concentrate on SETI for the time being, taking Gundolf's suggestion to clear my local preferences and see if I can get it to start sending me the reserve units I need.

I will report back on my progress.

---

Maybe I did not explore my options on Climate Prediction enough, because the work units I got took months on my 2.8 GHz machine which is idle 20 hours a day, or more. The deadlines were so tight that BOINC was always giving them priority, so I finally just took CP out of the picture.

At any rate, I am going to concentrate on SETI for the time being, taking Gundolf's suggestion to clear my local preferences and see if I can get it to start sending me the reserve units I need.

I will report back on my progress.

CPDN is not a good choice for a system that crunchs part time (especially on a single CPU system) unless it is the only project. BTW, AFAIK, CPDN doesn't have firm deadlines, but Boinc doesn't know this, as long as you trickle regularly the project will accept the results.

---

Boinc V7.2.42
Win7 i5 3.33G 4GB, GTX470

Looks as though things are starting to awake....

---

Official Abuser of Boinc Buttons...
And no good credit hound!

looks like all but the upload/download servers are up.

people will be getting work, but not downloading.

does boinc upload then report, or report than upload? criket says about 9.5 Mbits/sec.


well it did just a refresh its now 90 Mbits/sec guess every thing is running now.

Nick

---

looks like all but the upload/download servers are up.

people will be getting work, but not downloading.

does boinc upload then report, or report than upload? criket says about 9.5 Mbits/sec.

well it did just a refresh its now 90 Mbits/sec guess every thing is running now.

Nick

Upload first, then report completion.

And as usual since these 3-day outages started, there are limits in place to control download volume, although Jeff says the limits are higher this time.

And also remember that the Download Feeder process only holds 100 WUs at a time, and refills on a 5-6 second cycle, so even though there are several hundred thousand "Results Ready to Send", with all the empty hosts trying to refill caches, it may take "many" tries to download new WUs.

Here, patience is not just a virtue, it is a requirement.


Donald
Infernal Optimist / Submariner, retired