Posts by HAL9000


log in
1) Message boards : Number crunching : Panic Mode On (98) Server Problems? (Message 1697100)
Posted 14 hours ago by Profile HAL9000
You must remove obsolete app sections from app_info.xml. I still had the app section for AP6_win_x64_AVX_CPU_r2163.exe, and got the same message, after the outage.

I removed that app section and now all is well again.

You may remove the sections if you wish, but usually editing that file - unless you're experienced in how it all fits together - causes more problems than it cures.

The messages are harmless, if irritating - see previous answer.

Yes Richard, "may" would have been a better word to choose from me. I agree, that if you do not know exactly how it all fits together in app_info.xml, it's best to just ignore it, and let it be.

I imagine this is one of the reasons the removed the "Messages" tab and made the Event log so it took some effort to open. Fewer people seeing "something wrong"
2) Message boards : Number crunching : Bad new task initialization? (GPU crunching with ATI Radeon HD 4670, Windows 7 64bit) (Message 1697099)
Posted 14 hours ago by Profile HAL9000
Now I realize I posted in a wrong forum section; I think the present thread should be closed, and discussion should continue here:
http://setiathome.berkeley.edu/forum_thread.php?id=77631

Technically that is probably where the questions should be, but few people actually visit the Q&A section. You are more likely to get help with an issue in Number Crunching.
3) Message boards : Number crunching : Panic Mode On (98) Server Problems? (Message 1697069)
Posted 21 hours ago by Profile HAL9000
When was last time when AP Results ready to send was over 1,000? (before these new 2015 tapes)

Hey Grumpy Swede, are You happy now :)

When was the last time it was over 10?
4) Message boards : Number crunching : Panic Mode On (98) Server Problems? (Message 1696847)
Posted 1 day ago by Profile HAL9000
OMG, 2015 tapes? WOW!

Yes the big AP run on the 24th, was 2015 tapes too.

Twice within a 7 day period. Are they trying to spoil us or something?

Also noticed that someone did some cleanup on the tasks list. So it now only shows the currently active apps.
Application: All () · AstroPulse v7 () · SETI@home v7 ()
It is probably related to the updated Applications page which also now only shows the current apps.
5) Message boards : Number crunching : Tri-X R9 390X 8G D5 (Message 1696808)
Posted 1 day ago by Profile HAL9000
Until it passes through the x-ray machine ...

They have really ramped things up since the days I used to sell things on ebay.
6) Message boards : Number crunching : Panic Mode On (98) Server Problems? (Message 1696806)
Posted 1 day ago by Profile HAL9000
How do you thoroughly dry it?

Put it in a vacuum chamber.

Aquarium coupled to a Henry vacuum cleaner, you mean? :-)

Re-reading HAL's answer, I suppose he meant that he put the outside of the fan, the fan-blades and all that under the faucet, but not the actual axle housing. :)

I as talking about washing off a PCB not a fan, but I do sometimes wash my fans under the faucet. When I do that I run the water over the fan blades in one spot causing the fan to spin and basically wash itself. I wouldn't submerge one. The PCB in the fan could swell & separate.
7) Message boards : Number crunching : Tri-X R9 390X 8G D5 (Message 1696803)
Posted 1 day ago by Profile HAL9000
Plus of course the fact that one may get hit for import tax, which has to be paid before most carriers will release the goods :-(

Ummm... it's just like a 3kg package of socks. Yeah that's it!
8) Message boards : Number crunching : Tri-X R9 390X 8G D5 (Message 1696775)
Posted 1 day ago by Profile HAL9000
Sapphire Tri-X R9 390X 8G D5: are £350 in the UK :-( Plus of course £80-100 for a decent PSU to go with it.

That's about $550 USD. With them selling for $429 here I imagine shipping to the UK would cost less than $130... OK I just checked and it isn't that much less. $90 for 3-5 day shipping. There goes all my plans to makes lots of money selling GPUs to UK crunchers!
9) Message boards : Number crunching : "this computer has reached a limit on tasks in progress" (Message 1696772)
Posted 1 day ago by Profile HAL9000

The limits were imposed to prevent the db server dying. The limits have been enabled at two different time. With the 2nd time being a more or less permanent decision. I don't recall if it was before the first or before the 2nd time. There were ~11,000,000 results out in the field. Which at that point the db server comes to a grinding halt and everything crashes on a regular basis.



Perhaps it would not be such an issue if the deadline dates more closely matched the maximum number of days of work you can download. 56 days in some cases to do a work unit when you can only download 10 days plus and additional 10 days seems like overkill and adds substantially to the number of work units in the field.

Another possibility would be to limit the number of days of work you can download based on past performance instead of the number of work units based on absolute numbers. Three days total might cover most downtime events and chances are those that download the most work units for a three day period are the least likely to be the ones that download and run.

The long deadlines allow a broader range of machines to participate. So I don't see that changing soon. A dynamic deadline based on the performance & history of a machine could be interesting. That would require the BOINC devs to do some work that I imagine they don't see a need to do at this point.

The 2nd solution you mentioned is the basic idea of how the servers are suppose to work. That is where the "Max tasks per day" & "Consecutive valid tasks" in your host details are meant to come into play. In practice that system needs some work. As it ramps up really high & then reduces very little. One of my hosts has a "Max tasks per day" of nearly 10,000 & it does 90-120 tasks a day. If it glitched & started trashing work it wouldn't matter if it had a queue of 100 or 5000. It would still trash around 10,000 tasks. However that isn't the problem the queue limits are meant to resolve.
10) Message boards : Number crunching : Panic Mode On (98) Server Problems? (Message 1696762)
Posted 1 day ago by Profile HAL9000
No, nothing to panic about, but not really needed for me to start a new thread about this either. And if something should go awry with my cache, it's a handful of tasks only anyway.

Tomorrow my Areocool Areo-1000 case comes in, then I'll move my motherboard and innards over to the new case, take the HyperEVO 212 off my i5-2500K, bolt the Cooler Master Seidon 120v v2 on, and will go run some minutes of Passmark BurnInTest on it.

I ran that program on my i3-530 (with the same water cooler, and the fan on the radiator now reversed so it takes air over the radiator into the case, instead of blowing it over the radiator out of the case. :)), and even after an hour temperatures hadn't come above 45C. I've never seen that i3 so cold under 100% continuous load. So fingers crossed my i5 is going to take a step back in heat-spreading as well. It's idling at the moment at 44C and when doing some (light) load it peaks at 67C.

With temperatures the rest of the week being above 30C (86F) and on Saturday/Sunday expecting to be in the direction of 40C (104F), good cooling is required. Especially in a flat that warms up like an oven. Though I doubt I'll do much Seti during the next week / heat wave as I lack water cooling on the GPU.

The case comes with 3x120mm fans, I still have one 120mm spare fan lying on the table (although Twiggles, one of my cats, managed to puke her lunch on it. Complete with hairball. So I hope it's still all right, anyone got experience in this matter? ;-)), and the 120mm fan from the HyperEVO can probably be of use in there somewhere as well. Fingers crossed I have enough fan connectors on the motherboard.

I generally let the kitty vomit dry before attempting to clean it. I haven't had any issues with plastic surfaces similar to that of the composition of case fans. I did have a PCB that was puked on, but I was stripping parts off of it. So after sluffing off dried kitty puke I ran it under the faucet.
11) Message boards : Number crunching : AMD K6..... Just for fun:) (Message 1696753)
Posted 1 day ago by Profile HAL9000
They don't build them that well any more.What make is it?


I'd like to think it's AMD since they made many of the Intel 286-386 CPU's.. Truth is, as long as an item wasn't overclocked/overheated, it most likely still works today... I have a bunch of 386 CPU's that still work:)

The big killer of old old hardware was the barrel battery leaking and killing the PCB... As long as the mobo was properly maintained there is no reason for an old mobo to die... Ebay has plenty of supposedly working XT mobo's

Mine had Lithium 2025s or 2032s; I can't remember which... Probably the 2025s.

The li-ion CR3032 is the standard* motherboard battery used. CR2025 is more often found in automotive key fobs. However I think Zule is referring to the older 3 cell NiCd batteries that were soldered in.

*The good thing about standards is that there are so many to choose from!

Well; as I remember it, the SOYO 5EMA+ Motherboard for Socket 7, (AMD K6-2 350), used a different battery than what I had in my wristwatch. My wristwatch, at the time, used 2032s... Which is why I vaguely remember the Motherboards using 2025s; but, I could be wrong...


TL

Could be. That is why I added my not about standards. The CR2016 is a cheaper alternative & could have been used. They last about half as long as a 2032.
12) Message boards : Number crunching : Tri-X R9 390X 8G D5 (Message 1696751)
Posted 1 day ago by Profile HAL9000
I don't imagine either company will be having that issue again after the latest 3.5GB usable out of 4GB NV faced with one of their GPUs.

That was due to the memory configuration on the GTX 970, being split in one segment of 3.5GB and one of 512MB. The latter is only being addressed when a game or program actually tells it needs it.

Ah right, it was a speed issue not actually accessing it. I only keep tabs on so much info about NV hardware. Running 1080 it may not be an issue today, but cool kids are running 4K these days. Which does often uses >3.5GB of memory. I generally plan to use my GPUs for more than a year or however long the GPU releases cycles is. So something today may not be able to make full use of 8GB on the R9 390X, but it might soon enough. Currently the R3 390's only come in 8GB configurations. So even if you can ever use all of it there isn't another option.

The main question I come up with after looking at NVIDIA's response to the issue. Why does the 980 slow down so much in the same situations when it doesn't have restrictions accessing all of its memory? Either memory bandwidth isn't as important as they have been trying to tell us, or those titles are just not that memory intensive.
13) Message boards : Number crunching : AMD K6..... Just for fun:) (Message 1696735)
Posted 1 day ago by Profile HAL9000
They don't build them that well any more.What make is it?


I'd like to think it's AMD since they made many of the Intel 286-386 CPU's.. Truth is, as long as an item wasn't overclocked/overheated, it most likely still works today... I have a bunch of 386 CPU's that still work:)

The big killer of old old hardware was the barrel battery leaking and killing the PCB... As long as the mobo was properly maintained there is no reason for an old mobo to die... Ebay has plenty of supposedly working XT mobo's

Mine had Lithium 2025s or 2032s; I can't remember which... Probably the 2025s.

The li-ion CR3032 is the standard* motherboard battery used. CR2025 is more often found in automotive key fobs. However I think Zule is referring to the older 3 cell NiCd batteries that were soldered in.

*The good thing about standards is that there are so many to choose from!
14) Message boards : Number crunching : Tri-X R9 390X 8G D5 (Message 1696727)
Posted 1 day ago by Profile HAL9000
Just have a look at the power consumption-375W.

If you crunch with it 24/7 at (call it 300W) that will cost you about $500 or more per year in electricity alone.

The listed power consumption for those GPUs is less. Sapphire is listing the maximum power spec for the PCIe slot and power connectors for the cards. The actual spec for a 390X is 275W.
I think the general rule of thumb is that we normally see about 80% of the rated TDP running SETI@home apps. Which puts the power consumption closer to 220W. However the whole system would probably be looking at around 300w.
220w running 24/7 = 1927 KWh/yr, or 161 KWh/mo
300w running 24/7 = 2628 KWh/yr, or 219 KWh/mo
15) Message boards : Number crunching : Tri-X R9 390X 8G D5 (Message 1696723)
Posted 1 day ago by Profile HAL9000
I just wonder if the full 8GB can be used. And what kind of game you need for that. Although the new upcoming Deus Ex might be hitting it hard, nothing said about the likes of Doom 4, Hitman and maybe Fallout 4. Hope they don't expect me to play those with an Oculus Rift. :)

Of course, on the Fury X you pay for the high bandwidth memory, not so much for the water cooler. At 128GB/sec versus just 28GB/sec on GDDR5, that's no wonder.

I don't imagine either company will be having that issue again after the latest 3.5GB usable out of 4GB NV faced with one of their GPUs.

Once you start driving several monitors at 4K(AKA 2160p, UHD, or 3840 x 2160) resolution using all of that memory would not be hard to use. If they implemented DisplayPort v1.3 an 8K monitor could also be used.
Apparently GTA V on PC running 4K at 60 FPS is a good example of needing massively high end hardware. Actually running 4K at 30 FPS for GTA V requires some pretty hefty hardware too.
I just had to replace my monitor & my new one is only 1080. So I'm not worried about 4K gaming just yet. However when I replace my 61" TV I will be getting a 4K one. So I may still need to consider gaming on that.

For SETI@home use I don't think more that 2GB is needed running several instances. Maybe there could be some way for the apps to use more of the memory, if present, to speed things up in some way that no one has current thought to do.
16) Message boards : Number crunching : "this computer has reached a limit on tasks in progress" (Message 1696671)
Posted 2 days ago by Profile HAL9000
ok, then maybe a stupid question. Is there a way to fake more than 1 GPU? Can I make seti think I have 2 of the same GPU's.

I know you can setup a couple virtual boxes and double fake your computer so to speak but that is more work for me than worth.
99% of the time 100 task per works just fine. Anyways. it always is interesting to see how to circumvent the system.

Bob

The easiest way would be to modify the number of GPUs that that BOINC client reports to the servers. Which would require a code change. The source can be downloaded from the repository.
If you made BOINC think it had twice as many GPUs. This might get weird. An entry in an app_config.xml with something like <gpu_usage>2.0</gpu_usage> may need to be used. However BOINC may have other issues trying to address a nonexistent device.
Normally people just put in small cheap low end GPUs to bump their GPU count & then their main GPU runs through most of the queue. Mixing high and low end GPUs has its own set of problems as well.
17) Message boards : Number crunching : "this computer has reached a limit on tasks in progress" (Message 1696660)
Posted 2 days ago by Profile HAL9000
It seems to me that the limit on GPU tasks could be doubled without too much happening that would cause any problems. I know my GPU crunches at way more than twice the CPU rate. When a shutdown goes longer than expected a good portion of crunchers run out of GPU task way before the CPU tasks.

Is there someone who monitors this board that could implement a change on this. Assuming it would not be too difficult to code the change. I think we could triple it and not be problem but doubling to 200 would be a great improvement.

The only downside I see is the users that produces lot of invalids may increase the amount they are doing but I suspect it isn't a huge factor.

Just a thought.
Bob


People have been asking for the limits to increase since they were implemented. So you can guess how effective that is.
The change to the BOINC code was made to prevent systems with mixed GPU vendors from going into GPU starvation. That change lead to the limits change from 100 GPU tasks to 100 * GPUs per vender. Matt or Eric noted the change & decided to let it stand rather than clamp it back down. Then watch for any negative effects.
So now a system with 2 ATI, 2 Nvidia, & an Intel GPU could have 500 GPU tasks in total. If work was requested for all 3 types. When the change was first implemented it was a little broken. In that the server added all GPUs together & set the limit on tasks that way. So you could have those same 5 GPUs & only request work for Nvidia. Which would result in a limit of 500 instead of 200 tasks.
18) Message boards : Number crunching : "this computer has reached a limit on tasks in progress" (Message 1696644)
Posted 2 days ago by Profile HAL9000
There is a limit of 100 CPU task for 1 cpu chip despite however many cores it has

There is a 100 work unit limit for each GPU.

So for your system with 2 GPU and the 1 chip, the max is 300 work units

Why have the limit? It's a safe guard to make sure that someone doesn't download more work then they can possibly do.

What if someone downloaded 4000 work units then decided not to turn on his computer or if the computer had a melt down. Now all those work units are lost.

There is also a time limit as to when the work must be returned. Some computer are much slower than others. So sending them too many work units would result in work not completed in time.

Zalster

Edit..

If I remember correctly, this new system came about accidentally.

In the past it did learn your system and adjusted to that time to completion to match the request for how many days worth of work.

But I think they were trying to fix something and it morphed into what we have now. Others can give you a more detailed explanation.

The CPU limit is a flat limit of 100 CPU tasks regardless of the number of cores or CPU socks. I wish my dual socket servers were granted a limit of 100 tasks per socket like systems with multiple GPUs get.

The limits were imposed to prevent the db server dying. The limits have been enabled at two different time. With the 2nd time being a more or less permanent decision. I don't recall if it was before the first or before the 2nd time. There were ~11,000,000 results out in the field. Which at that point the db server comes to a grinding halt and everything crashes on a regular basis.

The first time the limits were put in place the limit of 50 tasks per CPU core & 400 total for GPU was used back in 2010-2011.

Many other projects also implement limits.
Moo! 250 tasks
PrimeGrid 100 tasks
Collatz 100 tasks
Milkyway has a more complex limit limit system. CPU cores * 3 up to a maximum of 48.
19) Message boards : Number crunching : Curious about cc_config.xml (Message 1696628)
Posted 2 days ago by Profile HAL9000
There have not been many additions to the cc_config.xml recently. When a new item is added the version where it first appeared is noted at the end of the entry.

However what has changed in there is now a GUI interface in BOINC manager to enabled or disabled the options. I would imagine if you use the GUI. That it writes all values to the file instead of just the items you select/change.
20) Message boards : Number crunching : Panic Mode On (98) Server Problems? (Message 1696622)
Posted 2 days ago by Profile HAL9000
HAL9000, look above ... it appears as though they may have ...

A bit strange this, had to redo one of my Parallella installations, reverted it to it's old host id's and seq_no plus one,
Seti Main sent new tasks, and didn't resend it's lost tasks, Seti Beta refused to resend the lost tasks and expired them, and issued new ones:

http://setiathome.berkeley.edu/results.php?hostid=7506529

http://setiweb.ssl.berkeley.edu/beta/results.php?hostid=74266

Claggy

It doesn't look like resend is enabled to me. As I have a machine that ended up fetching come MB GPU work just as I was shutting down BOINC to remove the MB GPU app. So it had about 40 tasks that are going to have to time out now. If resend was enabled those tasks would have been sent to the CPU by now.


Next 20

Copyright © 2015 University of California