[XPM] Noobproof VPS Primecoin All-In-One EZ Setup Script (Updated) - page 2.

gigawatt

full member

Activity: 168

Merit: 100

Good news! I've updated the script and now it works on Amazon EC2 instances.
Furthermore, primecoin launches as soon as the server reboots, so you'll always have it running.

nfuse

member

Activity: 97

Merit: 10

Thanks man for your work, even a noob like me got it working Wink

gigawatt

full member

Activity: 168

Merit: 100

Another tiny update.

The script now runs apt-get update before installing packages. This makes it work better with RamNode.
The start and stop scripts now use [[ -n "$(pidof primecoind)" ]] to check if primecoind is running. It should no longer kill itself on accident.

gigawatt

full member

Activity: 168

Merit: 100

Quote from: Qantaqa on August 10, 2013, 03:24:24 AM

I noticed the script still sets roundsievepercentage to 30 while hp9's default changed to 70, any specific reason for this, or can i just update this?

It's fixed now. HP9 came out after my last update.

Qantaqa

newbie

Activity: 39

Merit: 0

I noticed the script still sets roundsievepercentage to 30 while hp9's default changed to 70, any specific reason for this, or can i just update this?

gigawatt

full member

Activity: 168

Merit: 100

Quote from: Dumbo on August 08, 2013, 03:14:20 PM

Quote from: gigawatt on August 08, 2013, 10:50:37 AM

Correct.

Alternately, you can make a master wallet file on your computer, then copy it to each of the nodes. That way, you and all your nodes have the same wallets and you don't have to go fish addresses out of the nodes. If you go this route, be sure to stop primecoin, copy the wallet, then restart primecoin.

You can use WinSCP to transfer files to/from your nodes. Your login info will be the same as your PuTTY info.

Before I do something extremely stupid..

This is what I am doing.

1. From the regular primecoin client on my pc , I did Files>Backup Wallet and got the wallet.dat file
2. Then I used that wallet.dat file to replace .primecoin/wallet.dat file in the VPS using the ftp client.

That is what you meant right? Sorry for too many questions

Btw...if I have just cloned a droplet 3 more times, all 4 droplets have the same wallet.dat. Can I not just grab one of those wallet.dat files to Primecoin client on my PC?

You are correct on all counts. As long as all the nodes (including yourself) have the same wallet.dat file, they should stay in sync up to ~100 mined blocks.

I'd also recommend that you restart the haveged service. I noticed that it doesn't always run automatically after installation. You can accomplish this with:

Code:

service haveged restart

You can verify it's running with

Code:

ps aux | grep -v grep | grep haveged

Dumbo

member

Activity: 114

Merit: 10

Quote from: gigawatt on August 08, 2013, 10:50:37 AM

Correct.

Alternately, you can make a master wallet file on your computer, then copy it to each of the nodes. That way, you and all your nodes have the same wallets and you don't have to go fish addresses out of the nodes. If you go this route, be sure to stop primecoin, copy the wallet, then restart primecoin.

You can use WinSCP to transfer files to/from your nodes. Your login info will be the same as your PuTTY info.

Before I do something extremely stupid..

This is what I am doing.

1. From the regular primecoin client on my pc , I did Files>Backup Wallet and got the wallet.dat file
2. Then I used that wallet.dat file to replace .primecoin/wallet.dat file in the VPS using the ftp client.

That is what you meant right? Sorry for too many questions

Btw...if I have just cloned a droplet 3 more times, all 4 droplets have the same wallet.dat. Can I not just grab one of those wallet.dat files to Primecoin client on my PC?

Dumbo

member

Activity: 114

Merit: 10

Thanks! I will give it a try

gigawatt

full member

Activity: 168

Merit: 100

Quote from: Dumbo on August 08, 2013, 10:27:05 AM

So..I have to wait a bit before I can merge the accounts and get private keys from ./myinfo. Till a block is minted..right?

I just got got into bitcoins and cryptocurrencies, this is the first time I am mining (any cryptocurrency).

Correct.

Alternately, you can make a master wallet file on your computer, then copy it to each of the nodes. That way, you and all your nodes have the same wallets and you don't have to go fish addresses out of the nodes. If you go this route, be sure to stop primecoin, copy the wallet, then restart primecoin.

You can use WinSCP to transfer files to/from your nodes. Your login info will be the same as your PuTTY info.

Dumbo

member

Activity: 114

Merit: 10

So..I have to wait a bit before I can merge the accounts and get private keys from ./myinfo. Till a block is minted..right?

I just got got into bitcoins and cryptocurrencies, this is the first time I am mining (any cryptocurrency).

gigawatt

full member

Activity: 168

Merit: 100

Quote from: Dumbo on August 08, 2013, 07:06:53 AM

Please do not rip me apart, I am one of those complete noobs this guide was written for. I have started mining on a DigitalOcean Droplet. The primepersec is 528 (just in case anyone is wondering)

I have a few questions -

1. Using your tutorial on merging the primecoin accounts. I can open 3 other droplets, follow the same instruction to setup Primecoind and merge the - right?

2. The ypool droplet - do I have to keep the putty open? I saw someone complaining that they cannot mine as soon as they close the putty.

3. Amazon EC2 instances are the same thing as Digital Ocean Droplets right? will this work there too? I found this one some forum.. http://aws.amazon.com/big-data/powerof60/terms/ and was planning to use it.

4. If I signup for Linode , I would like to have a gigawatt's referral address. Since you took the time for us noobs, you deserve it.

Edit: I just noticed that his links are already referral links and that is totally fine by me. You deserve it sir!

Edit2: the command ./myinfo doesn't seem to work.

Thanks for enjoying the guide! I'll do what I can to help:

Yes, that's correct. I would recommend a different approach though. 1) Make one node, get it running and updated. 2) Shutdown the node and take a snapshot of the node. 3) Create a new node based on the snapshot. That way all the cloned nodes will have the same wallet files and should stay in sync up to ~100 mined blocks.
For the ypool one: no, you shouldn't have to. I have the script setup to run in the background, but the catch is that you can't "see" it work. You can only verify that it is doing work, not necessarily what it's accomplished so far. If you have doubts, run the start command in one session, connect with another PuTTY session, then check ps aux | grep miner.
Yes/no. They're similar, but as others have noted, Amazon EC2 instances require using sudo commands in certain places. If I recall correctly, it should be for each apt-get call. (So change apt-get install blahblah to sudo apt-get install blahblah)
I personally prefer Linode as well. $20 gets you an 8-core system vs $5 for a 1-core. The only drawback I've noted is that Linode doesn't let you "clone" nodes very easily, so you'll have to run the setup for each node.
As for ./myinfo: it only dumps the addresses that have a transaction associated with them. Transactions also include minting a block.

If you start having a handful of nodes to manage, I would recommend downloading multi-tabbed PuTTY. It'll let you opened tabbed session for each VPS and send commands to all of them at once.

Dumbo

member

Activity: 114

Merit: 10

Please do not rip me apart, I am one of those complete noobs this guide was written for. I have started mining on a DigitalOcean Droplet. The primepersec is 528 (just in case anyone is wondering)

I have a few questions -

1. Using your tutorial on merging the primecoin accounts. I can open 3 other droplets, follow the same instruction to setup Primecoind and merge the - right?

2. The ypool droplet - do I have to keep the putty open? I saw someone complaining that they cannot mine as soon as they close the putty.

3. Amazon EC2 instances are the same thing as Digital Ocean Droplets right? will this work there too? I found this one some forum.. http://aws.amazon.com/big-data/powerof60/terms/ and was planning to use it.

4. If I signup for Linode , I would like to have a gigawatt's referral address. Since you took the time for us noobs, you deserve it.

Edit: I just noticed that his links are already referral links and that is totally fine by me. You deserve it sir!

Edit2: the command ./myinfo doesn't seem to work.

gigawatt

full member

Activity: 168

Merit: 100

Wow, I just shot myself in the foot today.

I set up a new node to test out a ./notify-block script.
I had it enabled right from the beginning, but about 20 minutes in things were going really really slow.

Turns out that when primecoin was downloading the blockchain, it triggered the notification script about 90,000 times. Oops.

gigawatt

full member

Activity: 168

Merit: 100

Another round of updates!

Quote from: Pentel on August 02, 2013, 01:12:36 PM

Pic Here.
error: incorrect rpcuser or rpcpassword (authorization failed)

Usually that error is from editing your primecoin.conf file before stopping primecoin.
I've tweaked the ~/stop-primecoind script so it force kills it if it doesn't shutdown gracefully on it's own.

Pentel

full member

Activity: 172

Merit: 100

I still get error's but I think it's working:
Pic Here.

It my be working but ./myinfo does not work.
Error:
error: incorrect rpcuser or rpcpassword (authorization failed)

I wiped everything and started again, but with errors again.
See Pic God damn wrong OS!!!

It works now. Thanks Roll Eyes

lamiomni

newbie

Activity: 53

Merit: 0

Which settings are you using for a 1-CPU droplet @DigitalOcean?

Basic settings, on extra large instance (sieve size 1M, sieve percentage 10, round sieve size 30): 4.5k PPS, 550 5-chains/h

Current best tweaked settings (so far), on extra large instance (sieve size 250k, sieve percentage 3, round sieve size 45): 5.7k PPS, 840 5-chains/h

Looking for best 5-chains/h

Lower sieve sizes are better, on extra large instance (sieve percentage 10, round sieve size 30):
- 100k: 6.6k PPS, 660 5-chains/h
- 250k: 5.6k PPS, 650 5-chains/h
- 500k: 4.9k PPS, 650 5-chains/h
- 900k: 4.6k PPS, 600 5-chains/h
- 1.1M: 4.4k PPS, 540 5-chains/h

Lower sieve percentages are better, on extra large instance (sieve size 1M, round sieve size 30):
- 1: 7.1k PPS, 640 5-chains/h
- 2: 6.5k PPS, 600 5-chains/h
- 3: 5.6k PPS, 800 5-chains/h
- 4: 5.2k PPS, 670 5-chains/h
- 5: 5.1k PPS, 610 5-chains/h
- 15: 3.7k PPS, 510 5-chains/h

Higher round sieve sizes are better, on extra large instance (sieve size 1M, sieve percentage 10):
- 20: 4.1k PPS, 440 5-chains/h
- 25: 4.5k PPS, 550 5-chains/h
- 35: 4.4k PPS, 660 5-chains/h
- 40: 4.2k PPS, 730 5-chains/h
- 42: 4.1k PPS, 740 5-chains/h
- 45: 4.0k PPS, 735 5-chains/h
- 50: 3.8k PPS, 500 5-chains/h

gigawatt

full member

Activity: 168

Merit: 100

Quote from: lamiomni on August 01, 2013, 11:14:17 AM

Didn't try, the script is almost working out-of-the-box, you also need to "sudo make install" for gmp.

FYI, XPM priming (hp8, Ubuntu Server 13.04 64 bit) on Amazon EC2 with High-CPU On-Demand Instances:
- Medium (1.7 GiB of memory, 5 EC2 Compute Units): ~1,1kPPS
- Extra Large (7 GiB of memory, 20 EC2 Compute Units): ~4.5kPPS

I think that also may apply to the apt-get commands too.

I'd recommend running benchmarks on a few different parameter setups, you'd be surprised what kind of performance you can get.
On a 1 CPU (2.4GHz) DigitalOcean droplet, I can manage ~120 5-chain/hr or about ~1k PPS with optimized parameters.

Assuming that keep your debug.log file relatively clean (echo "" > ~/.primecoin/debug.log before you start), you can use this to calculate your running averages:

Code:

grep primemeter ~/.primecoin/debug.log | awk '{s1+=$4;s2+=$6;s3+=$8} END {printf"%10s: %10d\n%10s: %10d\n%10s: %10d\n","Prime/h",s1/NR,"Test/h",s2/NR,"5-Chain/h",s3/NR}'

lamiomni

newbie

Activity: 53

Merit: 0

Didn't try, the script is almost working out-of-the-box, you also need to "sudo make install" for gmp.

FYI, XPM priming (hp8, Ubuntu Server 13.04 64 bit) on Amazon EC2 with High-CPU On-Demand Instances:
- Medium (1.7 GiB of memory, 5 EC2 Compute Units): ~1,1kPPS
- Extra Large (7 GiB of memory, 20 EC2 Compute Units): ~4.5kPPS

I'm trying to tune the parameters a bit to find optimal settings.

gigawatt

full member

Activity: 168

Merit: 100

Quote from: lamiomni on July 30, 2013, 05:57:30 PM

For Amazon EC2 instances, you may be considering sudo for the apt-get and the cp commands and maybe more, since you are not root on them.

Absolutely true. On the Linode/DigitalOcean boxes, you are root.

I'm wondering if the install would work if you 1) saved it as a shell script and 2) ran the script with sudo. Most of it should work, but I'm not sure if primecoin would drop the config files where they should be.

lamiomni

newbie

Activity: 53

Merit: 0

For Amazon EC2 instances, you may be considering sudo for the apt-get and the cp commands and maybe more, since you are not root on them.

Code:

echo "Installing libraries..."
apt-get install build-essential curl git libboost-all-dev libdb++5.3-dev libdb5.3++-dev libminiupnpc-dev libssl-dev m4 -y

Code:

echo "Installing libraries..."
sudo apt-get install build-essential curl git libboost-all-dev libdb++5.3-dev libdb5.3++-dev libminiupnpc-dev libssl-dev m4 -y

Code:

cp primecoind /usr/local/bin/

Code:

sudo cp primecoind /usr/local/bin/

Topic: [XPM] Noobproof VPS Primecoin All-In-One EZ Setup Script (Updated) - page 2. (Read 17729 times)