BOINC Pentathlon 2020

[GrockleTD]

2 minutes ago, Eschew said:

Is this what you're looking for?

It's erm... On the first page of the F@H and BOINC Badge Request Thread...?

ah right. for some reason i thought that thread was for just F@H

[Windows7ge]

Lookie what I found in the trash at work. :3

Spoiler

If I'm lucky it just stopped working from overheating. It is extremely dusty.

[GrockleTD]

14 minutes ago, Windows7ge said:

Lookie what I found in the trash at work. :3

  Hide contents

If I'm lucky it just stopped working from overheating. It is extremely dusty.

aw man, jealous as hell. been wanting to get a switch for my personal space engineers server

[Windows7ge]

@Gorgon So I'm currently holding 1619WU's across my rigs. It seems the file sizes vary depending on the job but on one of my rigs they average out to be around 135MB/WU

 

I've got all my rigs back up and running. It does appear the boinc-client service was killing itself due to a lack of space as everything is running smoothly...for now.

 

@leadeater convincing me to write scripts was a great idea. This is beautiful:

 

The other script worked too, I was able to add all the VM's to the project in one fell swoop:

 

I have run into the same issue leadeater has on some of his rigs. Waiting for memory. I have more RAM that isn't being utilized the problem is they're 64GB RDIMM's. There's no way my desktops will accept them. Once the event starts and I switch everything to native it should be a less prevalent issue. I've adjusted the memory setting to allow up to 90% utilization all of the time. It's helped.

 

42 minutes ago, GrockleTD said:

aw man, jealous as hell. been wanting to get a switch for my personal space engineers server

Don't be jelly just yet. It could be dead dead, non-recoverable dead. I won't know until I have a go at the insides. If it is salvageable though it will be a great addition to the 25U rack that just got dropped off today. For all the basic 1Gbit networking needs I can connect all the servers to it then run one cable to the main switch instead of a dozen cables.

[leadeater]

1 minute ago, Windows7ge said:

@Gorgon So I'm currently holding 1619WU's across my rigs. It seems the file sizes vary depending on the job but on one of my rigs they average out to be around 135MB/WU

They get bigger as they run so average out only the active ones.

 

2 minutes ago, Windows7ge said:

So I'm currently holding 1619WU's across my rigs

I've got 6121 assigned to my account atm 

[Windows7ge]

11 minutes ago, leadeater said:

They get bigger as they run so average out only the active ones.

How do I check the file size of the active ones?

 

11 minutes ago, leadeater said:

I've got 6121 assigned to my account atm 

No comment.

[leadeater]

3 minutes ago, Windows7ge said:

How do I check the file size of the active ones?

No idea . I just noticed checking a few folders that the size changed a lot from getting the job and then at the 60% mark ish.

 

Total usage for me is 560GB using boincview.

[Windows7ge]

Just now, leadeater said:

No idea . I just noticed checking a few folders that the size changed a lot from getting the job and then at the 60% mark ish.

Doesn't matter to me so long as the room is available. I'll let them all go for a day as it does appear we can't bunker until jobs are due later than the 5th @ 00:00UTC which because of the side of the globe I'm on I believe will be the 4th @ 8:00PM. Once I see jobs with later due dates bunkering will commence.

[leadeater]

 

Wonder what happened here 

 

Spoiler

 

[Windows7ge]

3 minutes ago, leadeater said:

Wonder what happened here 

Yoouuu ran out of steroids?

[leadeater]

1 minute ago, Windows7ge said:

Yoouuu ran out of steroids?

Far more attractive woman came by, her name was Rosetta.

 

Spoiler

Just flexing on how much I was adding 

Really do love having these servers I can use, one day I'll build up my own army at home, one day.

 

[GrockleTD]

hmm. where is the BOINC stats tracker?

[Eschew]

4 minutes ago, GrockleTD said:

hmm. where is the BOINC stats tracker?

You can check the BOINC credits accumulated across all your BOINC projects on BOINCstats, after you register for an account there and set up your BAM! thingy. 

[Windows7ge]

1 minute ago, leadeater said:

Far more attractive woman came by, her name was Rosetta.

Better watch out I hear she's a RAM digger. That's all she wants you for.

[leadeater]

Added dual X5650 server, had to use Ubuntu Server 20.04, 18.04 weird grub issue on install I forget how to fix. Another one waiting to be done and I'll finish that tonight, last one needs CPU swap for same CPUs which I'll also get done tonight.

 

Think that taps me out for what I can get running.

[leadeater]

@Windows7ge

I thought of something rather dirty, anti bunker defense. Spin up a bunch of VMs and containers and claim all the tasks Rosetta has 3 days before the event, just grab as much as you can.

 

Is that cheating? I think it might be.... lol

[ffbr]

Sorry if it's already in other answer, super busy at the moment with a days old baby.

 

I got couple of idle instances that I can use for the Marathon, trying to find the best allocation in term of ram/disk space.

What's a reasonable estimate in term of ram need (GB per thread?). And if I spun them now in bunkering mode, what's a rough need in storage (same GB per thread)?

 

[leadeater]

2 hours ago, ffbr said:

Sorry if it's already in other answer, super busy at the moment with a days old baby.

 

I got couple of idle instances that I can use for the Marathon, trying to find the best allocation in term of ram/disk space.

What's a reasonable estimate in term of ram need (GB per thread?). And if I spun them now in bunkering mode, what's a rough need in storage (same GB per thread)?

1GB/thread with 1 or 2 GB left over, same goes for disk space.

[ffbr]

Another question concerning bunkering, I see that my rosetta task have a deadline that is before the start of the marathon. How does that works?

[leadeater]

15 minutes ago, ffbr said:

Another question concerning bunkering, I see that my rosetta task have a deadline that is before the start of the marathon. How does that works?

We can't until 3 days before, any sooner than that and your task will expire. I'm just letting mine run as normal then will limit upload at the 3 day mark.

[ffbr]

Ok, good to know.

At least I checked that my bunkering works, I limited both the upload rate and time frame allowed to 1 min per day.

I requested 5 days worth of task but somehow do not get that much on one of my computer. 

 

Gonna prepare the instance(s), we only have some ibm cloud available and I am not so use to it. 

[IntMD]

Just so everyone knows, with Rosetta, there is a target CPU run time in the account preferences, the default is 8 hours, but you can change this to make the units run longer or shorter depending on your circumstances from 2 hours to 1 day and 12 hours. The tasks are then supposed to end after that runtime has expired, and it has run it's last instance/checkpoint or something.

 

[Windows7ge]

After 1 full year this is what my WCG statistics look like:

By 5/1/2021 I will more than double these statistics and hopefully take over 1st place in the team.

 

8 hours ago, leadeater said:

@Windows7ge

I thought of something rather dirty, anti bunker defense. Spin up a bunch of VMs and containers and claim all the tasks Rosetta has 3 days before the event, just grab as much as you can.

 

Is that cheating? I think it might be.... lol

I think bunkering overall is cheating but everyone does it so it evens out. I get the feeling even if you grabbed as much work as you could with VMs it wouldn't starve the server of work to hand out.

[marknd59]

Can someone that knows what they are doing please confirm the following steps as to what I should be doing.

 

Tomorrow, May 2nd (sometime after 00:00 UTC), I should grab 3 days of Rosetta@home tasks then "suppend network activity". I then leave my systems crunching away until sometime after 00:00 UTC May 5th when I turn network activity back on to upload all the results I have stored up. I then carry on crunching  Rosetta@home tasks until someone tells me what the hell to do next.

 

Does that sound about right ?

 

Do we have an official team coach to tell us what we should be doing and when ?

[Gorgon]

Some of the Rosetta jobs are consuming over 2-2.5GB RAM:

  PID USER      PR  NI    VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND
 9848 boinc     39  19 2763276 2.498g  85540 R 100.0 10.6 207:13.16 rosetta_4.15_x8
10683 boinc     39  19 2323856 2.080g  85476 R 100.0  8.9 207:07.33 rosetta_4.15_x8
11089 boinc     39  19 2323440 2.079g  85536 R 100.0  8.9 207:04.17 rosetta_4.15_x8
 9505 boinc     39  19 2309228 2.066g  85524 R 100.0  8.8 207:16.04 rosetta_4.15_x8
11398 boinc     39  19 1663444 1.447g  82584 R 100.0  6.2 207:02.95 rosetta_4.15_x8
10206 boinc     39  19 1435428 1.228g  81420 R 100.0  5.2  19:37.80 rosetta_4.15_x8
13215 boinc     39  19 1067236 984232  80492 R  97.7  4.0   1:55.87 rosetta_4.15_x8
12106 boinc     39  19 1077072 929576  81580 R 100.0  3.8 206:51.54 rosetta_4.15_x8
13212 boinc     39  19  953784 870704  80580 R 100.0  3.5   1:58.33 rosetta_4.15_x8
12100 boinc     39  19  983052 836500  75500 R 100.0  3.4 206:56.61 rosetta_4.15_x8
11702 boinc     39  19  964000 817616  75564 R 100.0  3.3 206:59.53 rosetta_4.15_x8
 8474 boinc     39  19  678172 618280  88132 R  99.7  2.5 207:23.29 minirosetta_3.7
12102 boinc     39  19  614592 468192  75036 R  99.7  1.9 206:53.58 rosetta_4.15_x8
 1098 boinc     39  19  609132 464016  76064 R 100.0  1.9  75:51.06 rosetta_4.15_x8
10187 boinc     39  19  606576 461440  76040 R  99.0  1.9 207:10.81 rosetta_4.15_x8
12104 boinc     39  19  604340 459128  76060 R 100.0  1.9 206:53.59 rosetta_4.15_x8
 9036 boinc     39  19  600024 454796  76040 R  99.3  1.8 207:19.50 rosetta_4.15_x8
32673 boinc     39  19  599068 453840  76040 R 100.0  1.8  80:26.52 rosetta_4.15_x8
13378 boinc     39  19  300356 210152  69856 R 100.0  0.9   0:58.20 rosetta_4.15_x8

 

Eek!