SETI.Germany's 13th Annual Pentathlon (2022)

Gilthanis

[H]ard|DCer of the Year - 2014
Joined
Jan 29, 2006
Messages
8,516
At the moment I'm crunching on a bunker for Numberfields since my Universe bunker finished a couple hours ago. Now trying to turn in the Universe bunker and hopefully I'll be able to upload everything before the obstacle is over. Also hoping a bunch of the pending validation work units go through in the next 24 hours.

Sadly, my 1050Ti still hasn't been able to get even a single work unit on Einstein after several hours of trying so far. I absolutely hate having a GPU idle.
try making a bat file to hammer the server to grab work.

for /l %%x in (1, 1, 10000) do (echo %%x
"c:\Program Files\Boinc\boinccmd.exe" --project http://einstein.phys.uwm.edu/ update
ping 127.0.0.1 -n 180 > nul)

That is all you need to put in it and let it run. It will help for pulling a work to Einstein.
 

pututu

[H]ard DC'er of the Year 2021
Joined
Dec 27, 2015
Messages
2,360
For Einstein, because the task requires quorum 2, you want to download as many and as early as possible to improve validation rate.

For nvidia, I find optimum output by running 2 tasks per gpu. Running more tasks per gpu does not gain other than the need to allocate one thread per task which can be used for NumberFields or Universe.
 

SmokeRngs

[H]ard|DCer of the Month - April 2008
Joined
Aug 9, 2001
Messages
17,675
I tried the hammer for Einstein on my server (Linux) and no matter what it always tells me gui_rpc_auth.cfg is not found. I know the file is there or I wouldn't be able to run the client at all. I am using instances I have setup. I also tried it with the main client on a different project and got the exact same error.
 

motqalden

[H]ard|DCOTM x4
Joined
Jun 22, 2009
Messages
2,659
I tried the hammer for Einstein on my server (Linux) and no matter what it always tells me gui_rpc_auth.cfg is not found. I know the file is there or I wouldn't be able to run the client at all. I am using instances I have setup. I also tried it with the main client on a different project and got the exact same error.
Your log file in the client says that or you see that in the console after initiating the update? If they are instances you would need to make sure you use the right port number and password in the update command
 

SmokeRngs

[H]ard|DCer of the Month - April 2008
Joined
Aug 9, 2001
Messages
17,675
Your log file in the client says that or you see that in the console after initiating the update? If they are instances you would need to make sure you use the right port number and password in the update command
It's in the console when using the command. I triple checked the correct port number before doing anything and I'm using the exact same password for each instance.

I'll have to mess with it later as I'm busy the next couple of hours at least and won't be able to do anything.
 

SmokeRngs

[H]ard|DCer of the Month - April 2008
Joined
Aug 9, 2001
Messages
17,675
I think I figured out the problem.

I was using a command prompt to activate it and it was throwing that error. After I changed to the data directory of the instance and ran it the error is gone and it seems to be working.

Now to do the same in building a bunker to run for Universe so I have something to crunch after my Numberfields bunker is finished crunching.

Another note: If the command is run from the data directory of the instance there is no need to include --passwd Password so that can be skipped.
 
Last edited:

motqalden

[H]ard|DCOTM x4
Joined
Jun 22, 2009
Messages
2,659
I changed my faster gpus to 3 instances running 1 wu each and am doing much better
 

pututu

[H]ard DC'er of the Year 2021
Joined
Dec 27, 2015
Messages
2,360
SG and CNT are very close by in Einstein race. We are ahead of SG with one VII card only (~2.5M PPD). I'm sure someone in SG may be bunkering one VII card or more. That's all it takes.

1652732367448.png
 

DWolvin

2[H]4U
Joined
Nov 13, 2002
Messages
2,991
I'm still trying to get my lowly 1080ti to Einstein, so fat I've been able to create an account (I think), but I can't pull anything.
 

Gilthanis

[H]ard|DCer of the Year - 2014
Joined
Jan 29, 2006
Messages
8,516
I'm still trying to get my lowly 1080ti to Einstein, so fat I've been able to create an account (I think), but I can't pull anything.
Yeah...it is rough right now. All I can say is try running the scripts mentioned above to automate sending the "update" command to the project.
 

pututu

[H]ard DC'er of the Year 2021
Joined
Dec 27, 2015
Messages
2,360
I'm still trying to get my lowly 1080ti to Einstein, so fat I've been able to create an account (I think), but I can't pull anything.
One suggestion. Just run a few instances (say four) and hope one of them fills up some tasks and off you go and adjust the instances accordingly. This will only increase your probability to start crunching early and possibly load up some work. Some participants fire up more than what I suggested as you can see more consecutive BOINC computer ID number.
 

pututu

[H]ard DC'er of the Year 2021
Joined
Dec 27, 2015
Messages
2,360
Jeeper's 12th Day Commentary

"Does [H] gamble at the Javelin Throw or do they surrender?" What do you guys think we should do? ;)


Day 12 - 16 May 2022 - 12th Dose​


Ladies and gentlemen, a very special dose is waiting for you today. Extra strong with obstacle, garnished with a strong touch of javelin and a solid base of City Run. Feel the difference to yesterday, when an intense touch of Sprint was still integrated. Feel the euphoria, and the rush of happiness and fulfillment. Enjoy every drop of it, because I soon have to move on. Therefore, do not put forward tiredness or other excuses. It is time for today's dose.

Sprint (SRBase)
The final of the Sprint is done and caused one or two surprises in the end. Unfortunately, it was clear much too early that TAAT was going to win, but behind them, it became really exciting again. Silver goes to TSBT, who secured the second rank with a strong final day. This way, they also managed to fend off an attack by SG, who still made it to the bronze rang. Congratulations!

The victim of this was P3D, who have to settle for #4 in the end. The pace of SG and TSBT was too high. At least, CNT (#5) could not threaten the Green Planet any more. Instead, they had to watch out like hell for [H] (#6) achten, who came dangerously close. But the better end was in Czechia.

RKN crossed the finish line in their regular #7, quite clearly ahead of Ukraine, who will be very happy with #8. But it was close in the end, because AMD Users approached quickly. But in the end they could "only" reach #9, catching AF (#10) and EVGA (#11) on the final straight.

BOINC.Italy (#12) and TC (#13) followed entering the finish line quite relaxed. Behind them, TFFE (#14) managed to keep the narrow lead over Meisterkuehler (#15). Piece by piece, 2ch (#16), OCN (#17), BOINC@AUSTRALIA (#18), and LITOMYSL (#19) crossed the finish line.

In the end, Russia secured #20 and took the Confederation (#21) and Chinese Dream (#22) by surprise on the final day. UBT finished in #23 and could not take advantage of the weakness of the Chinese. BOINCstats (#24) and BOINC Synergy (#25) were not able to to this, either. Crystal Dream in #26 did not even start running and therefore finishes last.

Javelin Throw (NumberFields@home)
I will look more closely at the javelin tomorrow, as it is not clear yet for every team whether they throw with full power today. But it seems that TAAT secures gold and P3D bags silver.

The fight for bronze remains close, but the three-way fight might turn into a two-way fight. With their throw today, CNT forces SG to throw once again. Only this way, the Germans can win bronze on their own. On the other hand, [H] has not yet done anything today and would throw away any chance for bronze.

Ukraine (#7) confirms the very good throw from the previous attempt and could even outthrow AF (#6) with a similar final throw. In the end, the Ukrainians could also finish ahead of TSBT (#14), even though the Scots throw better again today.

TC (#13) appears to skip today's throw, as do BOINC@AUSTRALIA (#17), TFFE (#18), BC (#19), Meisterkuehler (#20), and UBT (#21). But the day is not over yet.

Obstacle Run (Universe@Home)
TAAT has now taken the lead also here and thus made optimal use of the bonus day. P3D (#2) cannot follow so far and now has to watch out for TSBT (#3), who are very fast today.

Ultimately, we will not see the full impact of the bonus day here until tomorrow. I assume that the teams mobilize all their reserves already now, and the final days will be rather quiet. Many preliminary decisions will be made today. Where should you take a closer look in the next few hours?

RKN (#6) and CNT (#7) are at odds with each other, currently with an advantage for the Germans. It could also get interesting between Meisterkuehler (#11) and OCN (#12), as the gap is still small.

BOINC.Italy (#13) and AMD Users (#14) have probably found each other at this Pentathlon, they are also fighting for the ranks here at the Obstacle Run. Behind them, the gaps are more calming. It gets really close in #24, where Crystal Dream has overtaken BOINCstats.

What is also true here is that the day is still long and the pitfalls of the track can quickly cause a surprise or two.

City Run (Einstein@Home)
At the top is currently TAAT ahead of P3D ahead of TSBT. Unfortunately, BOINC@AUSTRALIA (#4) had to let them go, so bronze is now far away. But at least the Koala group has placed itself clearly ahead of [H] (#5) and SG (#6).

CNT (#8) found out how important #7 is for RKN. OCN, on the other hand, is taking it easy. Behind them, TC sucessfully fought for #10 and roasted the francophone rooster in #11. Ukraine is currently in #12 and appears relatively safe from the chasers.

While EVGA (#13) and BOINC.Italy (#14) are dueling, 2ch (#15) and Chinese Dream (#16) are taking it easier. Do not hurt me and I will not hurt you, appears to be their guiding principle. The opposite is true behind them, as LITOMYSL and AMD Users fight for #17.

Everything is still possible when it comes to the ranks 19-23. BOINCstats and UBT are leading the pack, but cannot feel safe. BOINC Synergy (#21), TFFE (#22), and Meisterkuehler (#23) are virtually tied.

Overall Standings
Again, just a quick look today here, with the hope for a clearer picture tomorrow. TAAT and P3D continue to lead the field. Behind them, TSBT (#3) keeps SG (#4) at bay. It will be very difficult for the Germans to still wean the Scots from bronze.

CNT has moved up into #5. This was to be expected, as RKN (#6) lost many points at the Javelin Throw. [H] (#7) will probably overtake them as well, but has to throw well once more to do so. While Ukraine improved to #8, AF (#9) and TC (#10) dropped a bit.

The gaps are also close behind them, whether for #11 (BOINC.Italy, EVGA, OCN, AMD Users), for #16 (2ch, Chinese Dream), or for #21 (UBT, BC). Maybe we will know more here tomorrow.

The days of decision are upon us. Three disciplines are still running. Today we get more clarity at the Obstacle Run and the Javelin Throw.

  • Does [H] gamble at the Javelin Throw or do they surrender?
  • Can P3D fight back at the Obstacle Run?
  • Will the Australians keep #4 at the City Run?

You see it remains exciting at the BOINC Pentathlon.

Yours, Jeeper

[Based on stats at 11:00 UTC]
 

SmokeRngs

[H]ard|DCer of the Month - April 2008
Joined
Aug 9, 2001
Messages
17,675
I've been trying for a day to get some work units for a 1050Ti and haven't gotten a single one yet. That includes having the script running to keep up the attempts at work and it's been going for probably 15 hours now. At this point I don't expect to see any work on that card.

Luckily my RX570s have been able to keep a steady trickle of work units going. I suspect priority for work units is going to cards/clients which were previously crunching and active before Einstein got inundated with requests after the other GPU projects finished.

Sadly, while I've gotten some more work for Universe, new work is very sporadic and hardly any completed work has been uploaded back to the servers. I had a smallish bunker I started to send back yesterday soon after the obstacle went live and I'd be surprised if even half those work units have been sent back.
 

SmokeRngs

[H]ard|DCer of the Month - April 2008
Joined
Aug 9, 2001
Messages
17,675
Jeeper's 12th Day Commentary

"Does [H] gamble at the Javelin Throw or do they surrender?" What do you guys think we should do? ;)


Day 12 - 16 May 2022 - 12th Dose​


Ladies and gentlemen, a very special dose is waiting for you today. Extra strong with obstacle, garnished with a strong touch of javelin and a solid base of City Run. Feel the difference to yesterday, when an intense touch of Sprint was still integrated. Feel the euphoria, and the rush of happiness and fulfillment. Enjoy every drop of it, because I soon have to move on. Therefore, do not put forward tiredness or other excuses. It is time for today's dose.

Sprint (SRBase)
The final of the Sprint is done and caused one or two surprises in the end. Unfortunately, it was clear much too early that TAAT was going to win, but behind them, it became really exciting again. Silver goes to TSBT, who secured the second rank with a strong final day. This way, they also managed to fend off an attack by SG, who still made it to the bronze rang. Congratulations!

The victim of this was P3D, who have to settle for #4 in the end. The pace of SG and TSBT was too high. At least, CNT (#5) could not threaten the Green Planet any more. Instead, they had to watch out like hell for [H] (#6) achten, who came dangerously close. But the better end was in Czechia.

RKN crossed the finish line in their regular #7, quite clearly ahead of Ukraine, who will be very happy with #8. But it was close in the end, because AMD Users approached quickly. But in the end they could "only" reach #9, catching AF (#10) and EVGA (#11) on the final straight.

BOINC.Italy (#12) and TC (#13) followed entering the finish line quite relaxed. Behind them, TFFE (#14) managed to keep the narrow lead over Meisterkuehler (#15). Piece by piece, 2ch (#16), OCN (#17), BOINC@AUSTRALIA (#18), and LITOMYSL (#19) crossed the finish line.

In the end, Russia secured #20 and took the Confederation (#21) and Chinese Dream (#22) by surprise on the final day. UBT finished in #23 and could not take advantage of the weakness of the Chinese. BOINCstats (#24) and BOINC Synergy (#25) were not able to to this, either. Crystal Dream in #26 did not even start running and therefore finishes last.

Javelin Throw (NumberFields@home)
I will look more closely at the javelin tomorrow, as it is not clear yet for every team whether they throw with full power today. But it seems that TAAT secures gold and P3D bags silver.

The fight for bronze remains close, but the three-way fight might turn into a two-way fight. With their throw today, CNT forces SG to throw once again. Only this way, the Germans can win bronze on their own. On the other hand, [H] has not yet done anything today and would throw away any chance for bronze.

Ukraine (#7) confirms the very good throw from the previous attempt and could even outthrow AF (#6) with a similar final throw. In the end, the Ukrainians could also finish ahead of TSBT (#14), even though the Scots throw better again today.

TC (#13) appears to skip today's throw, as do BOINC@AUSTRALIA (#17), TFFE (#18), BC (#19), Meisterkuehler (#20), and UBT (#21). But the day is not over yet.

Obstacle Run (Universe@Home)
TAAT has now taken the lead also here and thus made optimal use of the bonus day. P3D (#2) cannot follow so far and now has to watch out for TSBT (#3), who are very fast today.

Ultimately, we will not see the full impact of the bonus day here until tomorrow. I assume that the teams mobilize all their reserves already now, and the final days will be rather quiet. Many preliminary decisions will be made today. Where should you take a closer look in the next few hours?

RKN (#6) and CNT (#7) are at odds with each other, currently with an advantage for the Germans. It could also get interesting between Meisterkuehler (#11) and OCN (#12), as the gap is still small.

BOINC.Italy (#13) and AMD Users (#14) have probably found each other at this Pentathlon, they are also fighting for the ranks here at the Obstacle Run. Behind them, the gaps are more calming. It gets really close in #24, where Crystal Dream has overtaken BOINCstats.

What is also true here is that the day is still long and the pitfalls of the track can quickly cause a surprise or two.

City Run (Einstein@Home)
At the top is currently TAAT ahead of P3D ahead of TSBT. Unfortunately, BOINC@AUSTRALIA (#4) had to let them go, so bronze is now far away. But at least the Koala group has placed itself clearly ahead of [H] (#5) and SG (#6).

CNT (#8) found out how important #7 is for RKN. OCN, on the other hand, is taking it easy. Behind them, TC sucessfully fought for #10 and roasted the francophone rooster in #11. Ukraine is currently in #12 and appears relatively safe from the chasers.

While EVGA (#13) and BOINC.Italy (#14) are dueling, 2ch (#15) and Chinese Dream (#16) are taking it easier. Do not hurt me and I will not hurt you, appears to be their guiding principle. The opposite is true behind them, as LITOMYSL and AMD Users fight for #17.

Everything is still possible when it comes to the ranks 19-23. BOINCstats and UBT are leading the pack, but cannot feel safe. BOINC Synergy (#21), TFFE (#22), and Meisterkuehler (#23) are virtually tied.

Overall Standings
Again, just a quick look today here, with the hope for a clearer picture tomorrow. TAAT and P3D continue to lead the field. Behind them, TSBT (#3) keeps SG (#4) at bay. It will be very difficult for the Germans to still wean the Scots from bronze.

CNT has moved up into #5. This was to be expected, as RKN (#6) lost many points at the Javelin Throw. [H] (#7) will probably overtake them as well, but has to throw well once more to do so. While Ukraine improved to #8, AF (#9) and TC (#10) dropped a bit.

The gaps are also close behind them, whether for #11 (BOINC.Italy, EVGA, OCN, AMD Users), for #16 (2ch, Chinese Dream), or for #21 (UBT, BC). Maybe we will know more here tomorrow.

The days of decision are upon us. Three disciplines are still running. Today we get more clarity at the Obstacle Run and the Javelin Throw.

  • Does [H] gamble at the Javelin Throw or do they surrender?
  • Can P3D fight back at the Obstacle Run?
  • Will the Australians keep #4 at the City Run?

You see it remains exciting at the BOINC Pentathlon.

Yours, Jeeper

[Based on stats at 11:00 UTC]
Considering I have a couple of small bunkers for the Javelin that are almost done and ready to be tossed, there damn well better not be any surrendering. Hah, as if that was possible. Besides, with the shit show that is Universe right now, there's no reason not to put some resources into Numberfields. Despite the small bunkers I'll probably end up running Numberfields while the Javelin is active just so my CPUs have something to do instead of wait on Universe work units.

It's a pity our first throw wasn't as strong as our second or we'd be in really good shape.
 

wareyore

Hard DC'er of the Month November 2021
Joined
Jan 1, 2014
Messages
787
Jeeper's 12th Day Commentary

"Does [H] gamble at the Javelin Throw or do they surrender?" What do you guys think we should do? ;)


Day 12 - 16 May 2022 - 12th Dose​


Ladies and gentlemen, a very special dose is waiting for you today. Extra strong with obstacle, garnished with a strong touch of javelin and a solid base of City Run. Feel the difference to yesterday, when an intense touch of Sprint was still integrated. Feel the euphoria, and the rush of happiness and fulfillment. Enjoy every drop of it, because I soon have to move on. Therefore, do not put forward tiredness or other excuses. It is time for today's dose.

Sprint (SRBase)
The final of the Sprint is done and caused one or two surprises in the end. Unfortunately, it was clear much too early that TAAT was going to win, but behind them, it became really exciting again. Silver goes to TSBT, who secured the second rank with a strong final day. This way, they also managed to fend off an attack by SG, who still made it to the bronze rang. Congratulations!

The victim of this was P3D, who have to settle for #4 in the end. The pace of SG and TSBT was too high. At least, CNT (#5) could not threaten the Green Planet any more. Instead, they had to watch out like hell for [H] (#6) achten, who came dangerously close. But the better end was in Czechia.

RKN crossed the finish line in their regular #7, quite clearly ahead of Ukraine, who will be very happy with #8. But it was close in the end, because AMD Users approached quickly. But in the end they could "only" reach #9, catching AF (#10) and EVGA (#11) on the final straight.

BOINC.Italy (#12) and TC (#13) followed entering the finish line quite relaxed. Behind them, TFFE (#14) managed to keep the narrow lead over Meisterkuehler (#15). Piece by piece, 2ch (#16), OCN (#17), BOINC@AUSTRALIA (#18), and LITOMYSL (#19) crossed the finish line.

In the end, Russia secured #20 and took the Confederation (#21) and Chinese Dream (#22) by surprise on the final day. UBT finished in #23 and could not take advantage of the weakness of the Chinese. BOINCstats (#24) and BOINC Synergy (#25) were not able to to this, either. Crystal Dream in #26 did not even start running and therefore finishes last.

Javelin Throw (NumberFields@home)
I will look more closely at the javelin tomorrow, as it is not clear yet for every team whether they throw with full power today. But it seems that TAAT secures gold and P3D bags silver.

The fight for bronze remains close, but the three-way fight might turn into a two-way fight. With their throw today, CNT forces SG to throw once again. Only this way, the Germans can win bronze on their own. On the other hand, [H] has not yet done anything today and would throw away any chance for bronze.

Ukraine (#7) confirms the very good throw from the previous attempt and could even outthrow AF (#6) with a similar final throw. In the end, the Ukrainians could also finish ahead of TSBT (#14), even though the Scots throw better again today.

TC (#13) appears to skip today's throw, as do BOINC@AUSTRALIA (#17), TFFE (#18), BC (#19), Meisterkuehler (#20), and UBT (#21). But the day is not over yet.

Obstacle Run (Universe@Home)
TAAT has now taken the lead also here and thus made optimal use of the bonus day. P3D (#2) cannot follow so far and now has to watch out for TSBT (#3), who are very fast today.

Ultimately, we will not see the full impact of the bonus day here until tomorrow. I assume that the teams mobilize all their reserves already now, and the final days will be rather quiet. Many preliminary decisions will be made today. Where should you take a closer look in the next few hours?

RKN (#6) and CNT (#7) are at odds with each other, currently with an advantage for the Germans. It could also get interesting between Meisterkuehler (#11) and OCN (#12), as the gap is still small.

BOINC.Italy (#13) and AMD Users (#14) have probably found each other at this Pentathlon, they are also fighting for the ranks here at the Obstacle Run. Behind them, the gaps are more calming. It gets really close in #24, where Crystal Dream has overtaken BOINCstats.

What is also true here is that the day is still long and the pitfalls of the track can quickly cause a surprise or two.

City Run (Einstein@Home)
At the top is currently TAAT ahead of P3D ahead of TSBT. Unfortunately, BOINC@AUSTRALIA (#4) had to let them go, so bronze is now far away. But at least the Koala group has placed itself clearly ahead of [H] (#5) and SG (#6).

CNT (#8) found out how important #7 is for RKN. OCN, on the other hand, is taking it easy. Behind them, TC sucessfully fought for #10 and roasted the francophone rooster in #11. Ukraine is currently in #12 and appears relatively safe from the chasers.

While EVGA (#13) and BOINC.Italy (#14) are dueling, 2ch (#15) and Chinese Dream (#16) are taking it easier. Do not hurt me and I will not hurt you, appears to be their guiding principle. The opposite is true behind them, as LITOMYSL and AMD Users fight for #17.

Everything is still possible when it comes to the ranks 19-23. BOINCstats and UBT are leading the pack, but cannot feel safe. BOINC Synergy (#21), TFFE (#22), and Meisterkuehler (#23) are virtually tied.

Overall Standings
Again, just a quick look today here, with the hope for a clearer picture tomorrow. TAAT and P3D continue to lead the field. Behind them, TSBT (#3) keeps SG (#4) at bay. It will be very difficult for the Germans to still wean the Scots from bronze.

CNT has moved up into #5. This was to be expected, as RKN (#6) lost many points at the Javelin Throw. [H] (#7) will probably overtake them as well, but has to throw well once more to do so. While Ukraine improved to #8, AF (#9) and TC (#10) dropped a bit.

The gaps are also close behind them, whether for #11 (BOINC.Italy, EVGA, OCN, AMD Users), for #16 (2ch, Chinese Dream), or for #21 (UBT, BC). Maybe we will know more here tomorrow.

The days of decision are upon us. Three disciplines are still running. Today we get more clarity at the Obstacle Run and the Javelin Throw.

  • Does [H] gamble at the Javelin Throw or do they surrender?
  • Can P3D fight back at the Obstacle Run?
  • Will the Australians keep #4 at the City Run?

You see it remains exciting at the BOINC Pentathlon.

Yours, Jeeper

[Based on stats at 11:00 UTC]


Tell everyone we surrender.

Then throw FTW.
 

DWolvin

2[H]4U
Joined
Nov 13, 2002
Messages
2,991
Awww yeah! I got an account made. Set #2..? Step #3...Profit!

Next year it would be good to remind noobs like me to go make accounts at all the sites just in case.
 

motqalden

[H]ard|DCOTM x4
Joined
Jun 22, 2009
Messages
2,659
Awww yeah! I got an account made. Set #2..? Step #3...Profit!

Next year it would be good to remind noobs like me to go make accounts at all the sites just in case.

at around this point of the pent I always lament my lack of planning and then come the next pent I forget to get prepared early. It's a vicious cycle lol.
 

bluestang

Limp Gawd
Joined
Dec 14, 2018
Messages
511
Ticklers were able to grab some E@H work overnight so I'm firing on all 4 cylinders (Turbo of course) :)

I think they fixed their server issue as well.
 

Gilthanis

[H]ard|DCer of the Year - 2014
Joined
Jan 29, 2006
Messages
8,516
Ticklers were able to grab some E@H work overnight so I'm firing on all 4 cylinders (Turbo of course) :)

I think they fixed their server issue as well.
Yeah...apparently their site had a bad config of some kind...
 

arnemetis

2[H]4U
Joined
Aug 2, 2004
Messages
3,921
My 5950x and 5800x have been dead in the water for two days, they just have a huge backlog of completed universe work units that wont upload. At this point they probably never will, and certainly not before the contest is over. Oddly enough my 5600x still has tons of work units in queue, but that too has a huge backlog of completed units that will not upload. Perhaps moving to linux was a mistake for the pent. So I entirely missed the 100% bonus. Really feels like this is less a contest of teams coming together to complete work and more about whatever team can configure the best scripts/exploits/hacks to force uploads and bypass the project backoff limits.

Perhaps for future contests the teams should get together and have a donation period beforehand, to have the selected projects upgrade their hardware & infrastructure to adequately process the work? I don't know, just trying to end on a less negative note, I feel defeated.
 

wareyore

Hard DC'er of the Month November 2021
Joined
Jan 1, 2014
Messages
787
My 5950x and 5800x have been dead in the water for two days, they just have a huge backlog of completed universe work units that wont upload. At this point they probably never will, and certainly not before the contest is over. Oddly enough my 5600x still has tons of work units in queue, but that too has a huge backlog of completed units that will not upload. Perhaps moving to linux was a mistake for the pent. So I entirely missed the 100% bonus. Really feels like this is less a contest of teams coming together to complete work and more about whatever team can configure the best scripts/exploits/hacks to force uploads and bypass the project backoff limits.

Perhaps for future contests the teams should get together and have a donation period beforehand, to have the selected projects upgrade their hardware & infrastructure to adequately process the work? I don't know, just trying to end on a less negative note, I feel defeated.
I'm in the same boat, on W10. So much U@H I'm going to dump if not uploaded by tomorrow.
 

pututu

[H]ard DC'er of the Year 2021
Joined
Dec 27, 2015
Messages
2,360
Jeeper's 13th Day Commentary

Let us make our last Javelin throw count. We have only one chance left ;)
Throw target: 3.4M

Day 13 - 17 May 2022 - 13th Can​


Ladies and gentlemen, please do not push. I know that time is running out, soon our cure for body and soul will be over. With today's combination of Obstacle Run and City Run, you can relax a little, before it comes to a climax tomorrow with a pinch of javelin again. Let yesterday's dose still resonate and take the dose for today.

Javelin Throw (NumberFields@home)
As expected, TAAT secured gold with the javelin and P3D nailed silver firmly to the tree. Both now do not need to do anything tomorrow and can focus their energy on the other two disciplines.

SG (#3), on the other hand, has to do extra time. The Germans made it exciting whether they would throw far enough yesterday. In the end, their javelin landed almost on the spot to edge out CNT (#4) and especially [H] (#16). However, [H] threw away the fight for the medal yesterday. Apparently they did not believe in it any more and so now can move on to the remaining disciplines tomorrow. SG and CNT have to do extra time, tomorrow's throw will decide here.

Expectedly, RKN (#5) had to say goodbye to the podium, but this will not grieve the Rechenkrafties. [H] will probably still pass them, but that is it. For AF, #6 is just a temporary stop, as Ukraine (#7), TSBT (#15), and [H] will probably still pass.

It is also still open who will take the final spot in the top 10. Chinese Dream (#8) has the best cards here, but has to do some work for it tomorrow. The reason for that is LITOMYSL (#24), who can make a giant leap with a good throw tomorrow, at least to #11.

Behind them, the field is a bit confusing. OCN (#9), EVGA (#10), AMD Users (#11), BOINC.Italy (#12), 2ch (#13), and TC (#14) are quite close together. We have to wait and see how the final day will go, who will throw at all, and how far.

BOINC@AUSTRALIA (#17) has skipped yesterday and now has to act tomorrow, as otherwise TFFE (#18) is threatening to pass. The Confederation did what they had to do yesterday and defended #19 for now. However, please do not forget that LITOMYSL will still overtake them.

This will probably also be true for Meisterkuehler. Currently in #20, they could still end up close to the top 10. UBT (#21) skipped yesterday and thus lost the chance to attack the Confederation. However, the Brits have nothing to lose any more. BOINCstats (#22) and BOINC Synergy (#23) are too far away. Russia is safely at #25, as is Crystal Dream at #26.

Obstacle Run (Universe@Home)
The last bonus day has probably clarified the battle lines. There were no major shifts. The question now remains whether the teams maintain their pace here or concentrate on the other disciplines.

Rarely does it get exciting at the top on the long track at this point of the race. However, this time this is the case. TAAT is still leading, but P3D is coming closer and closer. It remains to be seen whether the Green Planet can still strike here. TSBT (#3) is lagging behind a bit, it is hard to tell if they can still catch up.

SG is still in #4, but cannot feel too safe any more, because [H] (#5) is getting closer. RKN in #6 appears safe, as do CNT in #7 and TC in #8. AF (#9) has nothing to fear from behind, as they stumbled to a decent lead over Ukraine (#10).

Meisterkuehler will probably be satisfied with #11, although it is still unclear whether OCN (#12) can and will chance this. BOINC.Italy (#13) and AMD Users (#14) are also not too far behind, so this is not decided quite yet.

Unless Chinese Dream (#16) still ignites a rocket, EVGA should be safe at #15. Behind them, we find two duels that will probably last until the finish. TFFE (#17) and BOINC@AUSTRALIA (#18) on the one hand, and UBT (#19) and 2ch (#20) on the other hand are close together, so you should keep an eye here, my dear reader.

LITOMYSL (#21), BC (#22), and BOINC Synergy (#23) do not threaten each other at the moment, and there are no signs that this will change. The situation is different between Crystal Dream (#24) and BOINCstats (#25), where everything is still possible.

City Run (Einstein@Home)
Here, the teams are currently struggling with the pitfalls of the track. While it used to be rock-stable for many years, it is putting the participants to the test this year as well. TAAT continues to lead clearly ahead of P3D and TSBT. The chances that there will still be changes in this order are difficult, but not impossible.

BOINC@AUSTRALIA (#4) is now under pressure, because CNT (#5) is pushing hard and left [H] (#6), SG (#7), and RKN (#8) standing. Especially [H] and SG will surely fight back. It is very close between OCN (#9) and TC (#10), with AF (#11) also still within striking distance. Ukraine is running in #12 and managed to outrun EVGA (#13), BOINC.Italy (#14), and 2ch (#15). But this does not bother the trio at the moment, as they just have eyes for each other.

The same is true for LITOMYSL (#16), Chinese Dream (#17), and AMD Users (#18). And while you are plowing through the field with your binoculars, feel free to take a look at the battle ground behind them. It is still very close here, as well, and there can be all sorts of entertaining shifts: Meisterkuehler have recently taken #19 and are now trying to outrun BOINCstats (#20), UBT (#21), TFFE (#22), and BOINC Synergy (#23).

The fight for #24 is also worth a look. Russia (#24) currently has the best cards here, but can feel the breath of BC (#25) and Crystal Dream (#26).

Overall standings
We are in the final hours and some things are still unclear. Only the Javelin Throw tomorrow will iron out one or the other fuzziness in the rankings. While TAAT is on the verge of overall victory and silver is already smiling temptingly at P3D, bronze still has to find a new owner (or vise versa?).

TSBT (#3) has the best cards here, as they will probably still score at the Javelin Throw tomorrow, which makes it quite difficult for SG (#4) to use the minimal chance. CNT (#5) cannot get in there any more, but probably [H] (#7) can, as they can also make gains at the Javelin Throw tomorrow. This could lead to RKN (#6) dropping to their usual seventh rank.

The fight for #8 will probably be decided between Ukraine and AF, although TC (#10) still has at least a theoretical chance. Behind them, the gaps are still close and there will certainly be shifts. Currently hot pairings include BOINC.Italy and EVGA (for #11), OCN and AMD Users (for #13 or even also for #11?), Chinese Dream and Meisterkuehler (for #17), TFFE and LITOMYSL (for #19 or even more?), UBT and BC (for #21), and finally BOINC Synergy and Russia (for #24).

The stage is set. The teams are ready. The final is upon us. But how will it end?

  • Can P3D win the Obstacle Run?
  • Can CNT still catch the Australians at the City Run?
  • Who wins overall bronze?

You see it remains exciting at the BOINC Pentathlon.

Yours, Jeeper

[Based on stats at 11:00 UTC]
 

Gilthanis

[H]ard|DCer of the Year - 2014
Joined
Jan 29, 2006
Messages
8,516
My 5950x and 5800x have been dead in the water for two days, they just have a huge backlog of completed universe work units that wont upload. At this point they probably never will, and certainly not before the contest is over. Oddly enough my 5600x still has tons of work units in queue, but that too has a huge backlog of completed units that will not upload. Perhaps moving to linux was a mistake for the pent. So I entirely missed the 100% bonus. Really feels like this is less a contest of teams coming together to complete work and more about whatever team can configure the best scripts/exploits/hacks to force uploads and bypass the project backoff limits.

Perhaps for future contests the teams should get together and have a donation period beforehand, to have the selected projects upgrade their hardware & infrastructure to adequately process the work? I don't know, just trying to end on a less negative note, I feel defeated.
This sadly would not be as simple as you put it. If the project is backed by a University, then there is a strong possibility that ANY donations sent to it wouldn't necessarily go to the project nor the infrastructure. Typically what happens is that the money just goes to the University and gets spent on other worthless things including a pay raise for the President. Even if the project got the money, typically the infrastructure in place is the same for the whole University and allotted a portion of its capabilities. It would be unlikely that the infrastructure would actually see any noticeable bump.

For non-University projects, that may help a little. However, there are a ton of factors that every project would have to consider differently. Then you have the case of people just don't want to buy other people free hardware for a science that may never see the light of day or even get published. You will find a lot of projects are a hobby of someone's and ran on whatever hardware and internet they have at their disposal. One project I am aware of is ran on a Pi device. For normal day to day, it works. But for a competition, you would have to pretty much rebuild the entire project.

The problems we have now, is that even the best and most capable projects just aren't good enough any more. Residential hardware (and commercial) has grown at speeds that dwarfs networking capabilities. What worked in the early 00's is still pretty much the backbone of a lot of places to this day in regards to speeds. Not much has really improved. I mean, we have bank locations that still operate on shitty DSL connections that are slower than my cable internet. You would think we would have better infrastructure than we do. Factor in all the other moving parts to bandwidth bottlenecks like Universe had by uploading a bunch of tiny files because they don't want their server to have to unpack them.... and you get into a whole different issue. No matter how fast your internet is, it could be hampered by bull-headedness. I don't know what it takes to move those files back and forth once they receive them as I've never hosted a project. However, I would think a dedicated server doing nothing but unzipping files would be a lifesaver during events like this.
 

arnemetis

2[H]4U
Joined
Aug 2, 2004
Messages
3,921
This sadly would not be as simple as you put it. If the project is backed by a University, then there is a strong possibility that ANY donations sent to it wouldn't necessarily go to the project nor the infrastructure. Typically what happens is that the money just goes to the University and gets spent on other worthless things including a pay raise for the President. Even if the project got the money, typically the infrastructure in place is the same for the whole University and allotted a portion of its capabilities. It would be unlikely that the infrastructure would actually see any noticeable bump.

For non-University projects, that may help a little. However, there are a ton of factors that every project would have to consider differently. Then you have the case of people just don't want to buy other people free hardware for a science that may never see the light of day or even get published. You will find a lot of projects are a hobby of someone's and ran on whatever hardware and internet they have at their disposal. One project I am aware of is ran on a Pi device. For normal day to day, it works. But for a competition, you would have to pretty much rebuild the entire project.

The problems we have now, is that even the best and most capable projects just aren't good enough any more. Residential hardware (and commercial) has grown at speeds that dwarfs networking capabilities. What worked in the early 00's is still pretty much the backbone of a lot of places to this day in regards to speeds. Not much has really improved. I mean, we have bank locations that still operate on shitty DSL connections that are slower than my cable internet. You would think we would have better infrastructure than we do. Factor in all the other moving parts to bandwidth bottlenecks like Universe had by uploading a bunch of tiny files because they don't want their server to have to unpack them.... and you get into a whole different issue. No matter how fast your internet is, it could be hampered by bull-headedness. I don't know what it takes to move those files back and forth once they receive them as I've never hosted a project. However, I would think a dedicated server doing nothing but unzipping files would be a lifesaver during events like this.
I understand with the universities, that is a problem. But the community could address this too, by blacklisting projects that act like that.

Yeah I think the zipping and unzipping component might hit the nail on the head. Boinc would need an update, but for example every hour it still hasn't uploaded work, update the zip file with all the work that's been completed. As I was reading your comment I was thinking of compression being the solution, and you got there anyway.

Perhaps another consideration could be moving away from single projects per event type mentality. For example obstacle run could be Universe + 9 other projects combined, that could have similar rewards. This would be more work for the competition to track, but might distribute the workload and bandwidth constraints. Should help with them running out of work units too, you could just have it cycle to another project that has work queued up for you in your clients.

I also think there needs to be some serious work put into the exploits side too. It's been born out of these limitations you've spoken of, but significantly punishes those who do not run endless scripts and instances over the filthy casuals like me. The problem is there's no one interested in taking action to enforce or fix this. The only possible option is boinc's software, and I doubt they care enough to implement anything to stop it. My suggestion above about each contest being multiple projects seems the only way, certainly there will be peak performance projects that will get hammered, but at least people won't sit for 3 days with no work.
 

motqalden

[H]ard|DCOTM x4
Joined
Jun 22, 2009
Messages
2,659
IF World Community grid hadn't been down for server migrations it likely would have been picked for the Marathon and could have handled it with ease.
 

motqalden

[H]ard|DCOTM x4
Joined
Jun 22, 2009
Messages
2,659
We moved up to 4th in Einstein today which gives us the #5 spot overall. will have to see if we can hold it or if there is teams hiding bunkers... I'm sure it's going to be down to the wire!
 

Gilthanis

[H]ard|DCer of the Year - 2014
Joined
Jan 29, 2006
Messages
8,516
IF World Community grid hadn't been down for server migrations it likely would have been picked for the Marathon and could have handled it with ease.
Maybe.... SETI.Germany likes picking a specific sub project with WCG now for some dumb A** reason. Even if picked as the Obstacle course, they probably would have had issues since the WCG server limits the amount per project to send at any given time. We have seen that in the past.

arnemetis , the problem with multiple projects lumped together is that they all have different scoring mechanisms. You would certainly find further exploitation based on this. Plus, it would take away from the theme of the event. Pentathlon being based on Olympics and keeping everything related to 5. May, 5th, 5 projects... The bigger problem with these events is that there is little anyone outside of a project can do to really make them better. The projects rarely improve upon the BOINC server code. Most never contribute to it at all. So, unless you get the project involved with the event and doing some control mechanisms like PG does with their events...there is little SETI.Germany or any other hoster can really do to make it any different. As far as the BOINC devs go, they pretty much have their hands full as it is. And they have the attitude that every projects needs are different. So, any implementation they put in would have to be a "need" that benefits all of the community if there is any real hope for such an implementation to be added. Then, it gets put on an ever growing list of things that needs to be attended but that also means when someone has the time to develop. They do it on their free time.....

Edit: They have also started to make changes to the events each year to keep the challenge interesting and fun. Like the Obstacle course this year was new. Since we have a limited number of viable projects now, they didn't want it to become stale. Previously, when we have more viable options, they didn't allow projects to be chosen 2 years in a row. They had to roll with the punches. They then started doing things like picking sub projects if the project included those stats in the stats exports, tweaked the events like javelin throw and Obstacle run, and now allow projects to be chosen every year if needed.
 
Last edited:

Gilthanis

[H]ard|DCer of the Year - 2014
Joined
Jan 29, 2006
Messages
8,516
Don't forget the David Anderson effect as well with BOINC lol
Yeah...as in everything still has to be approved by him and he doesn't do any real development unless he is paid to do it...lol.
 

Gilthanis

[H]ard|DCer of the Year - 2014
Joined
Jan 29, 2006
Messages
8,516
He's too busy pimping Science United :)

It's on the homepage of BOINC lol

(yes, I have a little disdain for him)
I think quite a few that understand the evolution of BOINC have a little disdain for him...lol.
 

SmokeRngs

[H]ard|DCer of the Month - April 2008
Joined
Aug 9, 2001
Messages
17,675
We moved up to 4th in Einstein today which gives us the #5 spot overall. will have to see if we can hold it or if there is teams hiding bunkers... I'm sure it's going to be down to the wire!
The only problem with waiting until the last minute thinking bunkers are going to save you is that they are likely to give a pretty good boost to other teams as well when multiples of the work units are required to be turned in before credit is given. I've had a considerable increase in points production in the past day with Einstein because of that. As it is my two lowly GPUs still have around a million points in pending work units I've already sent in. I can't imagine what other people have sitting in pending

So, let everyone bring out their bunkers. It's only likely to help me and everyone else on the team.
 

pututu

[H]ard DC'er of the Year 2021
Joined
Dec 27, 2015
Messages
2,360
Jeeper's 14th Day Commentary

Another 2 hours left to the finale. Notably we are just behind SG in Einstein race. Don't let your guard down yet.
"Ranks 4-8 will be decided between CNT, [H], RKN, SG, and BOINC@AUSTRALIA. Here the final spurt will decide everything."

Day 14 - 18 May 2022 - 14th Dose​


Ladies and gentlemen, please do not push. There is enough for everyone! Our mixture of javelin, City Run, and obstacle once again ensures a full stimulation. You will feel the slightly stronger touch of City Run, whereas obstacle and javelin are less dominant, but still slowly unfold their effect. Enjoy this last dose so that you feel the positive after-effects from tomorrow.

Javelin Throw (NumberFields@home)
Here the decision is underway and it looks like the teams want to make everything clear early on. I only selectively look at this decision today, and will honor every rank again tomorrow.

SG has already secured #3 and relegated CNT to #4. The Germans have to work on their timing, as they have always needed four throws for years and thus waste valuable energy that they need elsewhere.

[H] also did what they had to for securing #5. Meanwhile, the fight for ranks 7-9 is still on. TSBT, Ukraine, and AF are still looking for a decision, with AF being only able to watch and Ukraine being able to get to the top on their own.

As expected, LITOMYSL (#11) and Meisterkuehler (#12) moved out of the basement with their throw today. Behind them, it is still close, although OCN (#13) and EVGA (#14) are making no effort to throw until now. AMD Users (#15) and BOINC.Italy (#16) could still take advantage of this.

It is also not clear yet between 2ch and TC who will take #17. Both are throwing today, but it remains to be seen who can improve. Behind them, TFFE (#19) have done everything they had to do today and now have to wait and see what BOINC@AUSTRALIA (#20) does. The Australians still have it in their hands to move up one rank.

City Run (Einstein@Home)
While the podium is probably clear (TAAT, P3D, TSBT), the ranks behind are still being fought over. Ranks 4-8 will be decided between CNT, [H], RKN, SG, and BOINC@AUSTRALIA. Here the final spurt will decide everything.

TC, OCN, and AF continue to fight for #9, with the Chinese currently in the best position. EVGA and BOINC.Italy still have to agree who will finish in #13. The same is true for 2ch and LITOMYSL and #15.

Chinese Dream is not completely safe in #17 yet, as the AMD Users are still close enough to burst dreams. UBT and BOINCstats are fighting for #20, as do BOINC Synergy and TFFE for #22 behind them. And who will finish in #24 will be decided between Russia, BC, and Crystal Dream.

Obstacle Run (Universe@Home)
It is the nature of this long discipline that the gaps are larger here at the end. The fact that particularly the race for the win is still open delights the heart of the spectator and raises the pulse of the involved participants to unimagined heights. TAAT or P3D, that is the question here. Who will bag gold? The next hours will decide. And let us not forget TSBT, as they are not quite out of the race, either.

Behind them, the ranks seem clear. Attacks are still possible, but should have only a limited size after the last bonus day. I therefore do not really expect any changes. Maybe in #19, currently occupied by UBT. 2ch (#20) could still do something here. Or in #24, where Crystal Dream is trying hard to fend off BOINCstats (#25).

Overall Standings
Only a short look here, as well (otherwise I have nothing more to do tomorrow). With the final attempt at the Javelin Throw, the Overall Standings have sorted themselves out as well. TAAT, P3D, and TSBT are about to climb the podium. While SG has to settle for #4, CNT (#5) and [H] (#6) are still fighting for rank 5.

Ukraine has a good chance to keep #8, as they still have room for improvement with the javelin. OCN (#13) and AMD Users (#14) are currently tied on points. This constellation is also found between LITOMYSL (#18) and Chinese Dream (#19), and promises excitement until the end.

We are in the final hours of the 13th BOINC Pentathlon. Not all the dice have been cast yet. We are ready for an exciting final phase that will also answer the final questions.

  • Who will win the Obstacle Run?
  • Will Ukraine take #7 at the Javelin Throw?
  • Who wins sheet metal at the City Run?

You see it remains exciting at the BOINC Pentathlon.

Yours, Jeeper

[Based on stats at 11:00 UTC]
 
Top