[XPM] [ANN] Primecoin High Performance | HP14 released!

[96redformula]

I have 50 vps running with chainsperday going from 1.18 to 0.9 and I find between 15 to 21 blocks per day.
They were not cloned so entropy is not a problem.

50 VPS with how many cores?  Digital Ocean?

[jimoc]

I have 50 vps running with chainsperday going from 1.18 to 0.9 and I find between 15 to 21 blocks per day.
They were not cloned so entropy is not a problem.

50 VPS with how many cores?  Digital Ocean?

If you have cloned instances running is there a way to fix the lack of entropy?
I've got 10 AWS instances running but not seen any blocks for 3 days and they were all launched from the same AMI.

[Tamis]

I have 50 vps running with chainsperday going from 1.18 to 0.9 and I find between 15 to 21 blocks per day.
They were not cloned so entropy is not a problem.

50 VPS with how many cores?  Digital Ocean?

If you have cloned instances running is there a way to fix the lack of entropy?
I've got 10 AWS instances running but not seen any blocks for 3 days and they were all launched from the same AMI.

Maybe you could read a few post above yours and get your answer ?

[paulthetafy]

Maybe you could read a few post above yours and get your answer ?

Yes, it would answer most of the questions in this thread.
People this lazy don't deserve answers.

IMO entropy is not the issue.  I've never seen an instance even close to running out of entropy (without havaged / rng-tools installed).  I was running hundreds of ec2 instances, all cloned from one, without issue.

[FoldingTime]

Here's a useful oneliner if you want to see exactlya sorted list of when your blocks were found and if you are not comfortable with reading times and dates as
seconds since 1970... UTC.  

Code:

primecoind listtransactions | grep blocktime |  sed -e "s/.* : //" -e "s/,//" | sort | xargs -n 1 -I '{}' date --date=@'{}'

Nice one liner  Thanks.

[jimoc]

I have 50 vps running with chainsperday going from 1.18 to 0.9 and I find between 15 to 21 blocks per day.
They were not cloned so entropy is not a problem.

50 VPS with how many cores?  Digital Ocean?

If you have cloned instances running is there a way to fix the lack of entropy?
I've got 10 AWS instances running but not seen any blocks for 3 days and they were all launched from the same AMI.

Maybe you could read a few post above yours and get your answer ?

Yes, I am an idiot

[matt4054]

Yes, I am an idiot

Nope, you are just lazy: functionally it's less of a problem than idiocy, but ethically it's worse

Just kidding, I think it's just hard to follow every thread (especially when your watchlist contains 50+ topics)

[Dsfyu]

Yes, I am an idiot

Nope, you are just lazy: functionally it's less of a problem than idiocy, but ethically it's worse

Just kidding, I think it's just hard to follow every thread (especially when your watchlist contains 50+ topics)

Yeah, it's really hard to follow a thread like this especially when dealing with outside issues like work/school/etc.

I feel like I'm going to be lazy asking this question since it probably has been answered somewhere in the thread, but I have two similar computers mining. Both have Phenom II processors and one is running windows 7 and the other is running ubuntu 13.04. Both computers are running hp9. The one running it in windows is getting 0.7 cpd and the Ubuntu one is getting 0.5 cpd despite the fact that they have phenom IIs with the same number of cores and the same clock speed. Could this be caused by the fact that one of these is a quad core chip (the 0.7 cpd cpu) and one is one of the 6 core ones with two cores locked (the 0.5 cpd chip)? I'm just trying to figure out what I should be doing to get the one with lower performance to match the other one.

[Krusher33]

Yes, I am an idiot

Nope, you are just lazy: functionally it's less of a problem than idiocy, but ethically it's worse

Just kidding, I think it's just hard to follow every thread (especially when your watchlist contains 50+ topics)

Yeah, it's really hard to follow a thread like this especially when dealing with outside issues like work/school/etc.

I feel like I'm going to be lazy asking this question since it probably has been answered somewhere in the thread, but I have two similar computers mining. Both have Phenom II processors and one is running windows 7 and the other is running ubuntu 13.04. Both computers are running hp9. The one running it in windows is getting 0.7 cpd and the Ubuntu one is getting 0.5 cpd despite the fact that they have phenom IIs with the same number of cores and the same clock speed. Could this be caused by the fact that one of these is a quad core chip (the 0.7 cpd cpu) and one is one of the 6 core ones with two cores locked (the 0.5 cpd chip)? I'm just trying to figure out what I should be doing to get the one with lower performance to match the other one.

Which chips are they?  It could be their cache sizes.

[Tamis]

I started haveged on all my 50 vps just to see if I would notice any change in performance or find more blocks, I was extremely surprised with the result !

I had a great start of the day at noon with 10 blocks already found, after having started the haveged service I left for work and let everything run for 6 hours.
I was quite surprised to notice I wasn't receiving any emails notifying me of blocks found and for a good reason, for the first since I started mining with 50 vps I haven't found a single block in 6 hours...

I have just stopped the haveged service on all vps.  Now this could be pure bad luck but 50 * 6 = 300 hours or 12.5 days, even my desktop computer solves a block in less time than that !

It seems haveged is responsible but how and why I have no clue. Could just be bad luck, well extremely bad luck...

[matt4054]

It seems haveged is responsible but how and why I have no clue. Could just be bad luck, well extremely bad luck...

Well, that's puzzling to say the least. Keep us posted about your results. Good... luck

[one4many]

I started haveged on all my 50 vps just to see if I would notice any change in performance or find more blocks, I was extremely surprised with the result !

I had a great start of the day at noon with 10 blocks already found, after having started the haveged service I left for work and let everything run for 6 hours.
I was quite surprised to notice I wasn't receiving any emails notifying me of blocks found and for a good reason, for the first since I started mining with 50 vps I haven't found a single block in 6 hours...

I have just stopped the haveged service on all vps.  Now this could be pure bad luck but 50 * 6 = 300 hours or 12.5 days, even my desktop computer solves a block in less time than that !

It seems haveged is responsible but how and why I have no clue. Could just be bad luck, well extremely bad luck...

Well ... to be honest ... I don't believe you. 

[Tamis]

Well ... to be honest ... I don't believe you.  

wtf ? why would I make up something like that ? what would be the purpose ?

I'm already pissed enough, spare me your nonsense !

[one4many]

wtf ? why would I make up something like that ? what would be the purpose ?
I'm already pissed enough, spare me your nonsense !

The reason is that I went from 6 blocks a day (two days ago) to zero (the last 24 hours). With no change at all.

As a matter of fact ... the only thing which changed was difficulty (rising). I guess mass VPS mining is coming to an end.
All the nice coins you mined, you are going to spend to "keep your hope" in mining more in the foreseeable future.
Probably until you lost (your current) mined plus. In the end you had a lot of work, but no gain.
And AWS or whoever your VPS provider is the only one which made any real money.

[Tamis]

Well I just found a block after having stopped the haveged service.  Now this, again, could just be due to luck (fucking variance).

Now regarding mining on vps in order to reply to my friend one4many.

here are my last days results :

05/08 : 20 blocks
06/08 : 19 blocks
07/08 : 16 blocks
08/08 : 20 blocks

I see no sudden changes in those stats related to difficulty and yet I just saw an ENORMOUS one after starting the haveged service.

I said from the start that variance could be responsible even if it's very unlikely... yet luck is often unlikely.

Regarding difficulty, it rose a lot slower than the past previous days so it's not difficulty related. And no it's not the 3 or 4 blocks I didn't mine this afternoon that are responsible.

[Entz]

Its easy enough to check if you are having entropy issues or not. watch cat /proc/sys/kernel/random/entropy_avail , if it drops below 100-200 and doesn't go back up you have a problem,

Make sure you are only using one of haveged or rng-tools as having both is likely not idea.

Some systems have hardware sources that can be used (with rng-tools ), on EC2 : medium power instances are fine (floats around 120), higher powered ones (i.e. cc2) used to hit 0 after a minute or so (which can cause the daemon to pause).  I have been meaning to try rng-tools but haven't had the chance.

In the beginning it used to burn through it so fast I pretty much needed it. Now the instances seem fine without it. Now, that being said there is no guarantee that havegd is generating good entropy (unique enough), so that may be the problem. Or just bad luck.  I had a pi find a block at difficulty 9, that is like 10pps.

I will disable mine and see what happens, cant hurt right

[one4many]

..
Now regarding mining on vps in order to reply to my friend one4many.
..

Who said I'm your friend? 
Anyhow ... let's see what your stats say tomorrow.

You could be as unlucky as I am 

[Tamis]

A question to the people that know linux better than me (that makes a lot of people) !

Do I need to reboot the machine to get a "non haveged state" after stopping the service or just stopping the service and restarting primecoind is be enough ?

[Entz]

A question to the people that know linux better than me (that makes a lot of people) !

Do I need to reboot the machine to get a "non haveged state" after stopping the service or just stopping the service and restarting primecoind is be enough ?

Just stopping the service is enough (or remove it if you like). No need to even restart primecoind

You can tell it isnt working if your entropy drops from around 3000, below 1000 (where haveged would kick in) and gets to around 120.

[Tamis]

Thanks Entz !

Everything around 130.

I'm so pissed I even tried this on all my vps... I pmed gigawatt about this since he recommends it with good faith I presume.