LTT Official Folding Month 2019!!!

[sanya567xxx]

3 hours ago, sanya567xxx said:

aight as I apparently can't connect to the server properly whatsoever, here's the zip of the WU if anyone wants free points or wtv.

https://drive.google.com/open?id=1zjia4eUYf_SK4vo2AE4SIuzLYlAfT7R6

please post a msg if you're able to do it, thx

should still be worth 100k+, shows 112k atm

did someone claim this finally? Am still unable to send this project. It's fully calculated so you only need to send it

[GOTSpectrum]

3 hours ago, justpoet said:

So…yes.  Electricity bills.  But, unless you're using a heat pump to heat your house, you're really just shifting cost from heating oil or gas towards electric to heat with, so the difference is still probably pretty negligible in the budget unless you have very high electric prices and very low fossil fuel prices.  If you use a heat pump (mini split, geothermal, etc) then you'll be using electricity either way, but the heat pump will use about 1/3 the amount to do the same heating job.

Here our hot water is electrically heat but Central heating is gas fired. The gas is about 1/4 the cost of the electric because we are 100% renewable. But, because I just have one machine in each room we use, it keeps the places we use warm and then we aren't Heating the entire house.

[Gorgon]

1 hour ago, Kochiha Ichihara said:

So what projects are we all working on at the moment?

http://blue.dwarf.ca/HFM/index.html

[vanished]

My control panel has been giving me a lot of BS about "bat state detected, recovering" etc. and even doing this enough times in a WU to give up.  Thing is, literally nothing has changed since the last week+ where everything was working perfectly so I guess there's no point pushing hard until that gets sorted out so I'm gonna throttle my system back for a while.

[Manmeyco]

1 hour ago, Kochiha Ichihara said:

So what projects are we all working on at the moment?

14227 on 1050 ti

14249 on Quadro 600

14228 on GT 635

13829 on Ryzen 3 1200

13830 on Pentium E2140

13828 on i5 4440

[Metallus97]

2 hours ago, Gorgon said:

I’ve really have to stop sleeping this month, which usually isn’t a problem with my insomnia but my new meds appear to be working.  Woke up to this:

stuck upload on one slot and the other slot in the same rig had a really slooooow upload and eventually recovered:

and I had an alarm set for 8:30 to wake me in the middle of my sleep cycle and slept right through it. Time to buy a new alarm clock ...

sexy graphs, where did you get that from?

[Gorgon]

2 minutes ago, Ryan_Vickers said:

My control panel has been giving me a lot of BS about "bat state detected, recovering" etc. and even doing this enough times in a WU to give up.  Thing is, literally nothing has changed since the last week+ where everything was working perfectly so I guess there's no point pushing hard until that gets sorted out so I'm gonna throttle my system back for a while.

I've seen this with the occasion WU that hits the GPU harder. The old OpenMM 22 Beta work units hit my RTX cards really bad and I had to lower the power limits and overclocks a lot but they would still eventually fail so I ended up just using Pascal Cards on the OpenMM 22 WUs

 

If your over-clocking reduce the clock by a step or two. If your not overclocking try either reducing the power limit or applying a negative clock offset.

[Metallus97]

I got an offer to rent a Vega 56 for the time of the event for 30$... I thinnnnnk AMMA DO ITT!

[Guest]

 

@Gorgon It just occurred to me that your results are more consistent because of your mix of cards vs. mine that suffer when they get the smaller WUs that take a while.

 

 

[Gorgon]

Just now, Metallus97 said:

sexy graphs, where did you get that from?

I have a vm running Zabbix and custom scripts that collect data.

 

It is essentially the output from

FAHClient --send-command queue-info

parsed and sent to the zabbix server as a "trapper" variable using a cronjob

[Metallus97]

Just now, Gorgon said:

I have a vm running Zabbix and custom scripts that collect data.

 

It is essentially the output from

FAHClient --send-command queue-info

parsed and sent to the zabbix server as a "trapper" variable using a cronjob

Dam dro! Nice I will give it a try! Did not know the command. THX

[Gorgon]

2 minutes ago, Den-Fi said:

 

@Gorgon It just occurred to me that your results are more consistent because of your mix of cards vs. mine that suffer when they get the smaller WUs that take a while.

 

 

I hadn't though of that

[Metallus97]

Who is interested in me writing a tutorial on how to use AWS, Google Cloud etc free trails to push some PPD?

I use this regularly for my research and yeh at the moment I am figuring out how to run FAH

 

[Gorgon]

3 minutes ago, Metallus97 said:

Dam dro! Nice I will give it a try! Did not know the command. THX

It is a PITA as the queue-info command does not order the output by slot index but by WU index so you have to use an array to enumerate the PyON output.

#!/bin/bash
# Fully Qualified Name of Zabbix Server
z_server='<server_name>'
#"Host Name" of Monitored Host (Must match that configured on Server)
z_host='<monitored_host_name>'

#Slot Points Per Day Array
declare -a slot_ppds

# The Slot Index (Starts at 0)
declare -i slot_idx

# The Number of slots
declare -i slot_num=0

# Get the Queue Info from the Local Folding Client and process it one
# line at a time.
FAHClient --send-command queue-info | grep '^  {' | {
    while read queue_line; do
        slot_idx=$(echo $queue_line | awk -F',' '{print $25}' | awk -F':' '{print $2}' | sed -e s/\"//g)
        slot_ppds[slot_idx]=$(echo $queue_line | awk -F',' '{print $12}' | awk -F':' '{print $2}' | sed -e s/\"//g)
        ((slot_num++))
    done
    let i=0
    while [ $i -lt $slot_num ]; do
        zabbix_sender -z $z_server -s $z_host -k gpu.ppd[$i] -o ${slot_ppds[$i]}
        ((i++))
    done
}
#Display ID for Enviroment
disp_id=0
#Xauthority File Location
x_auth=<path_to_xauth>
# This System has 2 RTX 2070 Super Hybrids, each with two queryable Fans, so we report fan:0 and fan:2
# ignoring the second fan of each
zabbix_sender -z $z_server -s $z_host -k gpu.fanrpm[0] -o $(DISPLAY=:$disp_id XAUTHORITY=$x_auth nvidia-settings -t -q [fan:0]/GPUCurrentFanSpeedRPM)
zabbix_sender -z $z_server -s $z_host -k gpu.fanrpm[1] -o $(DISPLAY=:$disp_id XAUTHORITY=$x_auth nvidia-settings -t -q [fan:2]/GPUCurrentFanSpeedRPM)

[sanya567xxx]

59 minutes ago, sanya567xxx said:

aight as I apparently can't connect to the server properly whatsoever, here's the zip of the WU if anyone wants free points or wtv.

https://drive.google.com/open?id=1zjia4eUYf_SK4vo2AE4SIuzLYlAfT7R6

please post a msg if you're able to do it, thx

should still be worth 100k+, shows 112k atm

the wu is 100% completed with results included but my client is not able to connect to neither the work nor collection server so this WU is sitting doing nothing last 6 hours

@Gorgon would you like to give this a try? i don't want the folding work to go to waste but it most likely will if nobody claims this.

 

[Guest]

2 minutes ago, Gorgon said:

I hadn't though of that

Yeah. I'm sitting here looking at 2080 Tis do 1.8M like hol up....

Kind of annoying, but hey...

Seems to happen less often in the single GPU systems.

[GOTSpectrum]

Well guys,

 

3 CT scans, an MRI, ECG, EEG and an ungodly number of blood tests we are home, feeling much better.

 

I'm very drained to say the least, but I've been keeping an eye on you lot and God damn have you been impressive.

 

I'm not sure if the blog will be written by me or @Gorgon but either way you will get an entry in the blog.

[Guest]

Just now, GOTSpectrum said:

Well guys,

 

3 CT scans, an MRI, ECG, EEG and an ungodly number of blood tests we are home, feeling much better.

 

I'm very drained to say the least, but I've been keeping an eye on you lot and God damn have you been impressive.

 

I'm not sure if the blog will be written by me or @Gorgon but either way you will get an entry in the blog.

Glad things are looking up.

Hope you get some much deserved rest.

[GOTSpectrum]

1 hour ago, Den-Fi said:

Glad things are looking up.

Hope you get some much deserved rest.

Rest is not for me, with a new born and being a carer to two people.

 

But to be honest I like it this way, it just isn't in my nature to rest. 

[Gorgon]

5 minutes ago, Den-Fi said:

Yeah. I'm sitting here looking at 2080 Tis do 1.8M like hol up....

Kind of annoying, but hey...

Seems to happen less often in the single GPU systems.

Which WUs are sucking the life out of the 2080 Tis? I'll pull my stats up on HfM and look at how they perform across my cards.

 

BTW - live HfM data here

[Gorgon]

Almost there folks:

448MPPD - Keep it up!

[Guest]

25 minutes ago, GOTSpectrum said:

Rest is not for me, with a new born and beyond a carer to two people.

 

But to be honest I like it this way, it just isn't in my nature to rest. 

Ah, a fellow ALL THE HOURS ARE MINE enthusiast.

[GOTSpectrum]

3 minutes ago, Den-Fi said:

Ah, a fellow ALL THE HOURS ARE MINE enthusiast.

That is true

[shaz2sxy]

ohman ive got both cards no on 14191

 

1070ti has 5 hours and 32mins...... wtf!!!

[GOTSpectrum]

2 minutes ago, shaz2sxy said:

ohman ive got both cards no on 14191

 

1070ti has 5 hours and 32mins...... wtf!!!

F