Topic: [XPM] Pool mining primecoin using DigitalOcean (VPS) - page 5. (Read 88579 times)

Question about Linux.. Server running Ubu 12.04 X64, and I want to copy all my files and wallet.dat and
everything that I currently have on one servers over to another servers. How would I be able to do that?

No, I do not use digital ocean. I cannot use snapshots.  What would you recommend? 

This...I have easily racked up five bucks in charges for instances that either never worked, or spent hours and hours hung executing processes. Kind of BS for a service that should have 99.9% uptime.

I have found the same problems and have arrived at the same conclusion. DO droplet cpu performance is spotty. Been running 15 for the last 36 hours or so, not a single block. I more than made my money back with early blocks but it no longer seems cost effective to run these things. To me at least.


Not worth it, imo. I tried a few, actually had one hang on boot (which in DigitalOcean terminology means 'Active', yep, they charged me the 2 hours it was stuck...), when finally running the 4 had completely different pps. One got @ 14K, the others between 6k and 10k. None of them hit their full potential. Ran them for a few hours, found a couple blocks but that was back a few days. I think the 8 core droplets are/were the best bang for the buck, you could pretty much get an even pps distribution through those ones.

I shutdown my droplets this morning.

Going to keep mining on only my home machines unless someone, cough*mikaelh, releases a SUPER cpu miner.  

CPU only coins I guess.. you can't compete with GPU's obviously

ok, cool. it was just an idea I was playing with. Another question, Can you use the DO services to mine for anything else?

I tried that on the first week and it was definitely not worth it.

Honestly, I have flipped my views on DO the more I play around with their service. It sort of blows. It seems like every time you spawn an instance, it's a total crapshoot as to whether you will get what is promised. I've also had a bunch of instances hang for ridiculous lengths of time on power off commands, and had two freeze altogether that I just told their engineer to destroy (shared wallet ftw).

They have been super chill about this onslaught of miners, and they deserve maaaaaaad credit for that, but it seems fucked to me that you could open two of the same instances, even in the same locale, and easily have one running ten times the speed of the other. Amsterdam is running about double the speed of the US servers right now for me. I know PPS is not the end-all indicator, but it's something, and I suspect something is seriously amiss with an 8-core instance runs slower than single cores for prolonged periods. I have seen it time and time again while playing with builds and resizing. I'd say maybe 20% of my droplets will just be fucked and hobbled upon creation, and I will have to try again and hope for a better roll of the die. I finally got some multi-core instances in Amsterdam that aren't totally crippled and showed performance increases that corresponded to the increase in cores. Hopefully it's a sign of improvement. Maybe we are just totally killing them, but then why the hell wouldn't they just kick some of the heavier abusers and stop letting their respectable customers suffer at our hands? I think it's likely they are revelling in the numbers to flash at their investors and suit dummies, and all the while just pushing lackluster service to all their users. Whatever. It's cheap and making us unsustainable money, I guess.


edit, forgot to answer the question:

As far as finding blocks...still seems to be a combination of computational power and luck. My PPS has increased over tenfold in the past maybe five days, and my rate of block finding hasn't, if that's an indicator. Some people are running way more than me and have been for a week and are still high and dry. Day two, I hopped on with only my laptop as I was out of town and was snatching a block every few hours. Slowly increased with the improved miners and snagging better instances, then I went from micro to 2 cpu instances in amsterdam, now sitting at like, 20k PPS and have stayed above a block/24hrs thus far.

If the 24-core instance ran as promised/equal to 24 micro instances, you could get maybe 30k pps...but who knows what that even means? Also, 24 micro instances would cost $.17/hr. Depends on how much you like to gamble, and how much you think XPM will be worth when you want to cash out. It would be a safer bet to buy some. Block reward is like, 12.2 something, my first block was 22. something.

ok, out of curiousity. I'f I were to lease one of their 24core, 96gb ram servers for 24hrs($33.85), would there be a point?  I don't know how much processing power  it takes to find primes at this point, but I'm thinking it might be an interesting experiment.

[/quote]


Something went wrong before this. Try pasting it line by line instead of all at once.


I'll try to make that section more clear.

When logged into the VPS enter:


If you have mined a block you will see something like:


Take the address from that output: "A***" and on the VPS enter:


The result of this is the private key used in the next step. "P*****"

Now from your main wallet:


It will then show in the list of transactions for your main wallet. If you're satisfied the transfer went properly you may remove it by entering into the VPS:

hi ya i had no issues until this point hope you understand this snippet 

root@PRESSEDHARD:~/primecoin/src# echo "rpcuser=******* rpcpassword=****************** gen=1"
rpcuser=******* rpcpassword=****************** gen=1
root@PRESSEDHARD:~/primecoin/src#
root@PRESSEDHARD:~/primecoin/src# > ~/.primecoin/primecoin.conf
root@PRESSEDHARD:~/primecoin/src# sudo mv primecoind /usr/local/bin/.
mv: cannot stat ΓÇÿprimecoindΓÇÖ: No such file or directory
root@PRESSEDHARD:~/primecoin/src# primecoind --daemon
primecoind: command not found
root@PRESSEDHARD:~/primecoin/src#


which one of these or how do you auto send to a wallet address as they are mined

Getting coins to another wallet:

There are a number of ways to do this:

After they've matured:

Code:

primecoind sendtoaddress yourhomeaddress amount


Before they've matured:

from vps: Is this digitalocean yes?

Code:

primecoind listtransactions # to find the receiving address I do this after?? long??
primecoind dumpprivkey $receiving_address where do i find "dumpprivkey"

from home:

Code:

primecoind importprivkey $privkey_from_vps


I then remove the wallet.dat from the vps to get all new keys, though its not necessary.
thanks for the help sounds like fun so i just have to have a play and the best thing is i am learning something new.
thanks
noel

ah okay. Thanks.

I think you see where I was going with it... because I wasn't sure if it uses the wallet to detect what is reaching up the blockchain. If that was the case, then we would be using them all together.

Thanks for clarifying that maco. Although the wallet has some overhead there is no dedicated miner at this point so each of your instances is going to have the full Primecoin blockchain. I expect this to change in a week or two, which I presume will make running a VPS miner much easier.

My point was that if in theory, I want to use 5 servers with 1 wallet, wouldn't that give me a leg up in mining a block because instead of it being 1 server at 1k pps with its own wallet, couldn't I just power level all servers together and earn more blocks this way?

Got a PM


My suggested method for moving coins around is in the OP. You dump the private key from the droplet and importprivkey into your main wallet.

Got my DigitalOcean stickers today!



You can request your own by emailing [email protected] .

I had been having trouble getting my clients to sync up from time to time so I added these nodes posted by Sunny (original thread).

Start trembling in fear.

So if you buy about 10 000 Raspberry Pi you might have something? lol

For those having issues with low amounts of RAM (i.e. 256 MB and/or 512 MB), I compiled it on my Raspberry Pi with a whopping 10 Primespersec (you're all going down!) just by changing the makefile.unix
There's one line with the compile flags, remove the "-O2" from it and it should be alright. The Pi is 512 MB of RAM so I know this works, but good luck to those with 256.
This also works for other "optimized" clients.

Seems the profitability of VPS mining is nearing an end.