Topic: [DVC]DevCoin - Official Thread - Moderated - page 298. (Read 1058949 times)

That coincidies with my own experience.

I am working on a side project, something like a coinmarketcap overview of about 50 coins. Mixed with api chartdata of the biggest exchanges.
Everything in one site.

My server has 24 gigs. And runs smoothly as long as I have about 20 daemons ... but if I try to start all 50 daemons (yes I compiled 50 daemons by hand  ) sooner or later everything crashes.

At one point even apache crashed and I couldn't access the server, had to restart.

Another strange thing I noticed: when I do a pidof most coins only list one process (which is good), but I have seen some daemons namely namecoind and infinitecoind list many more ids when I do pidof. Why do they need many processes instead of just one???

I basically need most daemons only to make a getblockcount call and getdifficulty of the respective coin. (additionaly to provide donation adresses for visitors, so they can donate to my project whatever coin they like)

The basic thing I want to do is calculate the total money supply for every coin. And I do not want to rely on external sources/blockexplorers because they could be offline (as I have seen happen many times with coinmarketcap.com) so I don't want to do it like them.

Do you know if there is a way to have a naked stripped down daemon who doesn't do anything else then provide geblockcount and moneysupply if available?

I have seen that some coins provide the variable "moneysupply" when you do a RPC call. Others don't and you therefor have to calculate the blockcount and then apply the halving rules over time etc to get to the amount of total actual coins in circulation.

PS I never had any problems with the devcoind daemon. I use version 32501, which is the newest one I hope.

Nowadays I have noticed Fedora defaults the swap partition to only a little larger than the RAM. Maybe for good reason, as it does seem to get into swap-thrashing if you do try the old twice the size of the RAM or the even older three times the size of the RAM for swap.

Basically you probably want a gigabyte per coin for long-established coins, and maybe even for the newfangled "almost as fast as GeistGeld" coins since they go through blocks fast; and a gig per merged mined coin.

Getting into swap while mining is a poor idea as it leads to p2pool complaining that daemons are taking more than 5 seconds to respond, and it might well be that it complains because slow access to getblock calls maybe degrades mining performance for all the coins it is merging.

-MarkM-

thanks. So I should try to somehow enlarge the swap partition. Is this possible?

Unix-type operating-systems use all available RAM for file buffers and such, so don't look at total RAM in use, look at the resident set and whether the swap partition is getting used up.

I0Coin used to need about four gigs though and GeistGeld would crash from lack of RAM on an 8 gig machine even no other coins running so I moved it to 16+ gig machines.

The new halfway-fixed GeistGeld (the quick-and-dirty fix) uses way the heck less RAM and the properly fixed I0Coin even less.

Bitcoin you can still run in a gig I think, but maybe not in half a gig or not run it well in half a gig anyway.

Most other coins, if not merged mined, seem to fit into half a gig or so.

-MarkM-

I was wondering:

Is there a way to limit the ram usage of a process in ubuntu server?

Is there a command for that?

I have installed a few daemons on a server, and they eat up all ram very fast.

Even the ancient devcoin code has yet to display the kind of massive RAM-usage that that GeistGeld and I0Coin did, because we only have one block per ten minutes on average whereas I0Coin is much faster thus has way the heck more blocks to deal with and GeistGeld is maybe even too darn fast to be practical so has even way the heck more than I0Coin.

Plus of course, GeistGeld and I0Coin might have more blocks that were mined using p2pool than DeVCoin does because they have spent more time not being on public pools thus having to be mined using p2pool by individuals instead of being mined maybe on a pool that does not put all the miner payouts into the coinbase transaction eating up RAM...

Also, one of the coders told me soem tiem back that he was in fact adapting the fixed I0Coin, not working directly from bitcoin code.

I had gotten the impression from the wording of his posts that he was ignoring the whole fixed I0Coin for some reason or for no apparent reason and was instead working from raw bitcoin code, but he corrected that impression.

That is part of why it seems weird to now hear that supposedly hs has backgtacked and is in fact working directly from bitcoin and creating a devcoin version that lacks the fixes that were put into I0Coin.

rsnel did the I0Coin fixes that is why we wanted to get rsnel if possible to do the other coins too, since he might grok his fixes/code better than others might.

-MarkM-

I didnt see any mem specific code rsnel wrote but I think bitcoin code had something if i can remember about writing to disk instead of using ram.. that was prob part of the fix and then other mem leaks getting cleaned up.. We wont know for sure until someone runs the daemon on the pool and logs the difference of memory usage over sample period of time.

It should be 4 shares for the qt images that is what I proposed.

Also do we have a bounty for ability to transact dvc via qr codes?

2 parts 1 ability to generate qr codes from your pc wallet. 2 we need a mobile wallet. Mobile wallets can use builtin scanners and mobile wallet would parse devcoin: {address} to send coins.

We can then add this to the webpage to promote devcoin.

If that is true, then the whole "merged mined coins use up more RAM and this is why" theory is maybe blown out of the water?

All the memory usage fixes for the merged mined coins were predicated on the theory that the massive coinbase transactions produced by p2pool get kept in RAM by the merged mined coins because the bitcoin chain's block against which they were merged is their proof of having been correctly merged mined, or something like that.

If that was not afterall what was realy going on, then maybe it was afterall a memory leak that caused GeistGeld and I0Coin to use massive amounts of RAM, not this business of the merged mining proof stuff in RAM...

Unless maybe the bitcoin code now somehow puts such proofs to disk like the memory-hog-fix does or something?

(And we care about that and should care about that not only because things that affect other merged coins might effect us but also because they are part of our merged mined coins family, providing a spread of different features using the same hashing power that secures our own chain. Also because we are all about free open source development and after bitcoin the other merged mined coins are our closest relations. We want merged mining to work nice and smooth for all the merged coins because we don't want merged mining to get a bad rep like "oh no you are merged mined, such and such a coin is merged mined and it has problems such and such and such and such...)

-MarkM-

There were already objections here: https://bitcointalksearch.org/topic/m.4095549

Unthinkingbit:The fee should remain a flat 1 dvc as far as I know, with no fork.

Bittzy:It can take a little while to add new writers to the receiver lists, but if still not on in a few days remind me and I'll chase it for you.

Giftculturewriting:Quite. As I said I think these things should be done properly, so taking one step at a time is important. There's some interesting discussion that would probably have been easier to follow if there was a separate devtome thread.

This address is from my computer and I checked it to be correct on your link.  Thank you, this is very exciting for me since its my first time.  On a side note, I sent Wiser a PM, since my last name begins with a "N", about setting up a writer's account, and have yet to hear anything.  I am sure the admins are extremely busy and just wanted to make sure that a delay is normal.  Thanks again!

He didnt do anything about low memory he just updated to bitcoin 0.8.5 and problem was gone.

If the admins choose for 3 shares for me, please divert 2 of the 3 shares to something else 

I am happy with getting anything  

Because I missed Ranlo's post, this bounty has been redone at:
https://bitcointalksearch.org/topic/m.4132653

Yep, that is where I found the link yesterday and downloaded it.

It was a very informative write up and the screen shots were very helpful.

I did a write up at Devtome: http://www.devtome.com/doku.php?id=testing_updated_devcoin_windows_client

The new client really behaves well. I love to see it linked via Devcoin.org once all small corrections have been made.

8 connections is the default max when your incoming port is not open.

So maybe your incoming port is not open.

Which might mean that your router does not support Plug and Play (or like mine it claims to but actually does not work with the plug and play library that bitcoin-derived cryptocoins use) or it might mean who-ever compiled the client compiled it without plug and play support or it might mean that who-ever compiled it compiled it with plug and play defaulting to disabled/off instead of defaulting to being enabled/on.

Then again it might also be that plug and play did work, your incoming port is enabled at your router, but it just hasn't happened to get any incoming connections yet.

(Its default max connections with an incoming port open is 125 I think.)

-MarkM-

I uploaded it earlier this evening and I also was impressed with the ease of use.  The visual aspect of the client was more like what I am used to and the installation was so smooth, which is important to technically challenged persons like myself .  It is still in the process of receiving the blocks, but I'm sure that will not be a problem.  Consistently, it is connected to 8 active connections, I'm not sure if that is the max amount or not.


DVC wallet:    13mtCyZTnMJRJw9nt7xqbiGU7tXrvkeAJB

I want Giftculturewriting to be an admin, and he chose to help categorize articles, so he do that with Weisoq and Wiser.