bathrobehero
Legendary
 Offline
Activity: 2002
Merit: 1051
ICO? Not even once.
|
 |
October 12, 2014, 03:58:05 AM |
|
Strange. It should read everything....what is your stagger atm?
1024 |
Not your keys, not your coins!
|
|
|
|
|
|
|
|
|
|
"Bitcoin: the cutting edge of begging technology." -- Giraffe.BTC
|
|
|
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
|
|
|
|
Irontiga
|
|
October 12, 2014, 04:05:00 AM |
|
Strange. It should read everything....what is your stagger atm?
1024 Use a vm...one running on a usb stick or something... |
|
|
|
|
bathrobehero
Legendary
Offline
Activity: 2002
Merit: 1051
ICO? Not even once.
|
|
October 12, 2014, 04:15:27 AM |
|
Strange. It should read everything....what is your stagger atm?
1024 Use a vm...one running on a usb stick or something... It won't make a difference because it seems it doesn't matter which application reads the plots, be it a miner, a file viewer or chkdsk, the result seem to be the same which is win7 choking on it's own file reading cache which I can't control. |
Not your keys, not your coins!
|
|
|
|
Irontiga
|
|
October 12, 2014, 04:16:38 AM |
|
Strange. It should read everything....what is your stagger atm?
1024 Use a vm...one running on a usb stick or something... It won't make a difference because it seems it doesn't matter which application reads the plots, be it a miner, a file viewer or chkdsk, the result seem to be the same which is win7 choking on it's own file reading cache which I can't control. Mine when you're not busy on your comp,and then when youre busy stop  |
|
|
|
|
bathrobehero
Legendary
Offline
Activity: 2002
Merit: 1051
ICO? Not even once.
|
|
October 12, 2014, 04:32:01 AM |
|
Mine when you're not busy on your comp,and then when youre busy stop My main rig is always busy and while it has more memory I can't have a loose app eating all the memory over time on it. The plots are on one of my gpu mining rigs. I stopped gpu mining while I was fooling around with plotting so that didn't interfere either. It's not even about profit anymore, just bugs the crap out of me that I can't figure it out  |
Not your keys, not your coins!
|
|
|
|
|
|
Blago
|
|
October 12, 2014, 06:54:56 AM |
|
I offer 1k BURST to someone who can tell me how can I (pool)mine with 5.5 TB without running out of 8 GB memory in a few minutes. I used all kinds of stagger sizes from 1024 to 16k, plot sizes from 2 GB to 5.5 TB using bipben's gpu plotter and tried uray's burst-miner and blago & dcct's miner. Using win7 and tried disabling prefetch/superfetch and various windows services without success.
It's Win7 problem https://bitcointalk.org/index.php?topic=731923.msg9086411#msg9086411 |
Relax, I’m russian!...
BURST-B2LU-SGCZ-NYVS-HZEPK
|
|
|
HoldTheLine
Member
Offline
Activity: 112
Merit: 10
|
|
October 12, 2014, 07:06:54 AM |
|
Who is the piece of shit trying to manipulate the graphs on Bittrex? Someone dumps 250,000 Burst and this dickhead uses his bot to make it look like someone bought 250,000 by manipulating the candle stick. This is the sorta shit that scares people away from buying into a coin - they aren't stupid and trying to fiddle the graphs is a sure fire way to loose potential investors.
|
|
|
|
|
|
uray
|
|
October 12, 2014, 07:31:15 AM |
|
...
What more do you want? BURST already has more built in features than any online banking out there.
this: The Burst dev needs to convert Burst network into decentralized dropbox ASAP, when users would pay for distributed encrypted storing of their files with BURST coins. This is the only way for Burst to survive.
Have you looked to this : http://storj.io/ ?? what i hope the difference between burst data storage and storj.io is we are using pay by data usage instead of storj's pay for disk space, so storing small files would be relatively very cheap, this way the burst node network can evolve into CDN serving nodes other than file storage the hardest thing to design that is that you have to rely that there are some data copies available whenever you want to access them. the best would be to be able to book a certain amount of bandwidth plus a guaranteed period of time plus a specified number of copies whenever you put a storage object into the network. compared to dropbox this can be realized by folder attributes. this concept is a service and no mining except if you add a asset which can only be earned if you put a specific nonce into the storage objects. maybe if you "host" 256mb encrypted data you get allowed to add one nonce for it. these nonces have to have another weight than simple mining nonces. each time you host the file as long as 10% of the booked time you are allowed to add another nonce or the weight of yours get adjusted (similar to the timekoin generator reward). if the period of booked time for the data package has exceeded it can either get renewed and if not the accounts hosted the data can keep their nonces for that package in reverse order (10 month hosted, 20 month additional reward). technically it is like a huge private torrent network backed by the blockchain (as tracker) and miners. for me it makes most sense at the moment to add filesharing like an asset. if you want to store files you require burst to buy this asset. if you only want to host data you receive this asset and can trade it for burst. this concept utilizes the mining infastructure people have (except the bandwidth usage) to store files but also requires people to pay similar like if the miners use their infastructure for pure mining. if i say 1 gb local storage may mine 18 burst within a month and i want to have 500 copies of my data i would have to pay 9000 burst for 1 gb cloud storage at the current diff. if you reduce the number of copies to 50 1gb would cost 50 dollarcent at current prices. if the price stays stable at the current level this calculation would be fine. if the price settles at 2000 satoshi or more anytime soon its a way too expensive to use. one of the most tricky designs is to make it for the miners similar provitable to host files as to mine. tricky are the redundant copies you require. nobody finally really cares to not loose data blocks on system failures. to prevent and recover this i can think of a infastructure which maintains itself. lets say you have defined periods of time to resync your hosted files or someone else gets them assigned and payed for. also the bandwidth usage is huge. this means if someone mines today with 4 tb he could only upload on average 4gb a day and blocks his line completely. if you factor this in you come to a ratio of how many files can be stored. with todays 4500 miners 18 tb cloudstorage could be realized. if you say you keep 100 copies of the files you end up at about 200gb which every miner stores. if you try to keep everything in correlation with the block rewards and market prices its hard to predict what will happen to such a service. i think over time the redundant copies make it much more expensive to use than if you host your data on a dedicated server. i currently dont know how price structures for data hosting look like but backblaze offers a complete backup for exactly 5 dollar a month. if i sum everything up this is what a dropbox like burst extension requires: -you define the number of copies you want to have -you define the guaranteed bandwidth you can access your files with -you define how long the files shall be stored -after the defined time the storage service renews itself based on current market prices for the same time if your balance allows it -you define who can access it -miners have to have similar profits wether they store files or use their diskspace for mining (correlation between block reward and diff) -the network reassigns lost storage slots to other people if a wallet could not be accessed for a certain period of time -storage may not cost more than regular cloud storage (depends on the number of copies) Rather than trying ensure a certain number of copies, an alternate approach is to make it most profitable to host data pieces that have the fewest other mirrors. Here's what I'm thinking: Use payments for storage as a 'storage reward'. The storage reward for a file will come around seemingly randomly but selected deterministically by running network state(gensig, etc) through a hashing-based algorithm that also takes a probability of a block having a storage reward for that file. When the storage reward time for a file comes up, there will also be a deterministically selected portion of the file. Miners who have the file stored can calculate a proof-of-knowledge of that piece of the file based off a merkle tree root that was stored in the blockchain when the file was stored, and announce an intent to collect the reward containing a hash of something along the lines of sha256(userId concat sha256(knowledge proof)) so that the knowledge proof is not yet revealed. After a set amount of time, any node can announce the proof-of-knowlege. A selection method is then used to select one of the miners to receive the storage reward. Using this method one shouldn't have to attempt to ensure a certain number of copies, as the most effective method of mining is browse the dht seeking out files that have the fewest number of copies available, and host copies of them, since they offer the best chance of receiving a storage reward, as they will have the fewest amount of people competing for the reward. it seems you are using mix of term file and chunk, shouldn't we use Piece (as in chunk in torrent file) instead of file, what is pros and cons of file or chunk ? |
|
|
|
|
|
uray
|
|
October 12, 2014, 07:36:39 AM |
|
...
What more do you want? BURST already has more built in features than any online banking out there.
this: The Burst dev needs to convert Burst network into decentralized dropbox ASAP, when users would pay for distributed encrypted storing of their files with BURST coins. This is the only way for Burst to survive.
Have you looked to this : http://storj.io/ ?? Please people, the file storage thing is purely BAD NEWS. All it'll do is make big data centers more money. Right now data centers don't have gpu's to plot, or on their servers with lots of hdd's good cpu's. This prevents them from wrecking burst. Now most people at home have a few TB's and a GPU or at least a good CPU. Please, the file storage thing will KILL burst. Leave that to storj, and if its that critical to you, well just leave and buy lots of storj. Please the file storage will not kill the burst , don't worry ! I've only told you that there is a simple service , a decentralized storage ( called storj) . If burst will come like storj ( without lose the PoC algorithm) I think will attract more users and users . it would not kill burst, but proof of capacity has potential to kill burst, 8 Petabytes is very small for datacenter, if they want to kill burst, they can setup a miner than hold 90% mining power, and dump all the mined burst, we need atleast total network capacity is larger than all datacenter in the world combined (just like bitcoin) as for file storage, yes datacenter will make most money out of it but it also strengthen the burst network and burst value, since all storage service data center would love to use burst protocol to provide their services |
|
|
|
|
bathrobehero
Legendary
Offline
Activity: 2002
Merit: 1051
ICO? Not even once.
|
|
October 12, 2014, 07:49:52 AM |
|
I offer 1k BURST to someone who can tell me how can I (pool)mine with 5.5 TB without running out of 8 GB memory in a few minutes. I used all kinds of stagger sizes from 1024 to 16k, plot sizes from 2 GB to 5.5 TB using bipben's gpu plotter and tried uray's burst-miner and blago & dcct's miner. Using win7 and tried disabling prefetch/superfetch and various windows services without success.
It's Win7 problem https://bitcointalk.org/index.php?topic=731923.msg9086411#msg9086411Mine doesn't clear the memory, it crashes. Should I try a huge staggersize like 800k? |
Not your keys, not your coins!
|
|
|
|
burstcoin (OP)
|
|
October 12, 2014, 07:53:37 AM |
|
...
What more do you want? BURST already has more built in features than any online banking out there.
this: The Burst dev needs to convert Burst network into decentralized dropbox ASAP, when users would pay for distributed encrypted storing of their files with BURST coins. This is the only way for Burst to survive.
Have you looked to this : http://storj.io/ ?? what i hope the difference between burst data storage and storj.io is we are using pay by data usage instead of storj's pay for disk space, so storing small files would be relatively very cheap, this way the burst node network can evolve into CDN serving nodes other than file storage the hardest thing to design that is that you have to rely that there are some data copies available whenever you want to access them. the best would be to be able to book a certain amount of bandwidth plus a guaranteed period of time plus a specified number of copies whenever you put a storage object into the network. compared to dropbox this can be realized by folder attributes. this concept is a service and no mining except if you add a asset which can only be earned if you put a specific nonce into the storage objects. maybe if you "host" 256mb encrypted data you get allowed to add one nonce for it. these nonces have to have another weight than simple mining nonces. each time you host the file as long as 10% of the booked time you are allowed to add another nonce or the weight of yours get adjusted (similar to the timekoin generator reward). if the period of booked time for the data package has exceeded it can either get renewed and if not the accounts hosted the data can keep their nonces for that package in reverse order (10 month hosted, 20 month additional reward). technically it is like a huge private torrent network backed by the blockchain (as tracker) and miners. for me it makes most sense at the moment to add filesharing like an asset. if you want to store files you require burst to buy this asset. if you only want to host data you receive this asset and can trade it for burst. this concept utilizes the mining infastructure people have (except the bandwidth usage) to store files but also requires people to pay similar like if the miners use their infastructure for pure mining. if i say 1 gb local storage may mine 18 burst within a month and i want to have 500 copies of my data i would have to pay 9000 burst for 1 gb cloud storage at the current diff. if you reduce the number of copies to 50 1gb would cost 50 dollarcent at current prices. if the price stays stable at the current level this calculation would be fine. if the price settles at 2000 satoshi or more anytime soon its a way too expensive to use. one of the most tricky designs is to make it for the miners similar provitable to host files as to mine. tricky are the redundant copies you require. nobody finally really cares to not loose data blocks on system failures. to prevent and recover this i can think of a infastructure which maintains itself. lets say you have defined periods of time to resync your hosted files or someone else gets them assigned and payed for. also the bandwidth usage is huge. this means if someone mines today with 4 tb he could only upload on average 4gb a day and blocks his line completely. if you factor this in you come to a ratio of how many files can be stored. with todays 4500 miners 18 tb cloudstorage could be realized. if you say you keep 100 copies of the files you end up at about 200gb which every miner stores. if you try to keep everything in correlation with the block rewards and market prices its hard to predict what will happen to such a service. i think over time the redundant copies make it much more expensive to use than if you host your data on a dedicated server. i currently dont know how price structures for data hosting look like but backblaze offers a complete backup for exactly 5 dollar a month. if i sum everything up this is what a dropbox like burst extension requires: -you define the number of copies you want to have -you define the guaranteed bandwidth you can access your files with -you define how long the files shall be stored -after the defined time the storage service renews itself based on current market prices for the same time if your balance allows it -you define who can access it -miners have to have similar profits wether they store files or use their diskspace for mining (correlation between block reward and diff) -the network reassigns lost storage slots to other people if a wallet could not be accessed for a certain period of time -storage may not cost more than regular cloud storage (depends on the number of copies) Rather than trying ensure a certain number of copies, an alternate approach is to make it most profitable to host data pieces that have the fewest other mirrors. Here's what I'm thinking: Use payments for storage as a 'storage reward'. The storage reward for a file will come around seemingly randomly but selected deterministically by running network state(gensig, etc) through a hashing-based algorithm that also takes a probability of a block having a storage reward for that file. When the storage reward time for a file comes up, there will also be a deterministically selected portion of the file. Miners who have the file stored can calculate a proof-of-knowledge of that piece of the file based off a merkle tree root that was stored in the blockchain when the file was stored, and announce an intent to collect the reward containing a hash of something along the lines of sha256(userId concat sha256(knowledge proof)) so that the knowledge proof is not yet revealed. After a set amount of time, any node can announce the proof-of-knowlege. A selection method is then used to select one of the miners to receive the storage reward. Using this method one shouldn't have to attempt to ensure a certain number of copies, as the most effective method of mining is browse the dht seeking out files that have the fewest number of copies available, and host copies of them, since they offer the best chance of receiving a storage reward, as they will have the fewest amount of people competing for the reward. it seems you are using mix of term file and chunk, shouldn't we use Piece (as in chunk in torrent file) instead of file, what is pros and cons of file or chunk ? Chunks are most likely the best way. With chunks user can download different chunks of the same file from different sources, hosts have better protection against knowing what they're storing, and keeping fixed size chunks makes it easier to ensure that the size of all possible proofs is greater than the size of storing the file itself |
BURST-QHCJ-9HB5-PTGC-5Q8J9
|
|
|
|
fanepatent
|
|
October 12, 2014, 10:50:03 AM |
|
The pool is up and the new ISP seems to work fine now. http://mine.burstcoin.info is in need of miners. We have 6 Tb mining directly into the asset account to increase share value. 4 more Tb on the way. Skype and email support available. |
BURST - BURST-58XP-63WY-XSVQ-ASG9A
|
|
|
go6ooo1212
Legendary
Offline
Activity: 1512
Merit: 1000
quarkchain.io
|
|
October 12, 2014, 01:10:40 PM |
|
The pool is up and the new ISP seems to work fine now. http://mine.burstcoin.info is in need of miners. We have 6 Tb mining directly into the asset account to increase share value. 4 more Tb on the way. Skype and email support available. I like the thing with skype support, Maybe I'll return mining there with one account after some plot optimisation... |
|
|
|
|
Patek
Member
Offline
Activity: 64
Merit: 10
|
|
October 12, 2014, 02:04:38 PM |
|
I know, its been asked before but hell, impossible to find anything in here...
Would mining with plots stored on usb 2.0 external hdd work out? Or is the transfer rate of 2.0 to slow to efficiently mine?
|
|
|
|
|
smaxer
Newbie
Offline
Activity: 47
Merit: 0
|
|
October 12, 2014, 02:28:40 PM |
|
I know, its been asked before but hell, impossible to find anything in here...
Would mining with plots stored on usb 2.0 external hdd work out? Or is the transfer rate of 2.0 to slow to efficiently mine?
USB 2.0 is too slow. |
|
|
|
|
|
fanepatent
|
|
October 12, 2014, 02:30:48 PM |
|
I know, its been asked before but hell, impossible to find anything in here...
Would mining with plots stored on usb 2.0 external hdd work out? Or is the transfer rate of 2.0 to slow to efficiently mine?
USB 2.0 is too slow. Its not exactly too slow. It will work but you might miss some deadlines on fast blocks. |
BURST - BURST-58XP-63WY-XSVQ-ASG9A
|
|
|
Patek
Member
Offline
Activity: 64
Merit: 10
|
|
October 12, 2014, 03:10:25 PM |
|
I know, its been asked before but hell, impossible to find anything in here...
Would mining with plots stored on usb 2.0 external hdd work out? Or is the transfer rate of 2.0 to slow to efficiently mine?
USB 2.0 is too slow. Its not exactly too slow. It will work but you might miss some deadlines on fast blocks. Would I on avarage hit at least 80% of blocks? |
|
|
|
|
smaxer
Newbie
Offline
Activity: 47
Merit: 0
|
|
October 12, 2014, 03:35:39 PM |
|
You have to read 1 GB for a 4 TB plot file. So it depends on how fast your USB device is and how big your plot file is.
|
|
|
|
|
Nevril
Member
Offline
Activity: 108
Merit: 10
|
|
October 12, 2014, 04:29:36 PM |
|
With USB2 and 4TB, you have 1GB to read @ 20MBps more or less... so ~50 sec (it should be 40-45 since 4TB aren't effective and if your device is good you can read up to 24-25 MBps)
On the positive side reading is completely overlapped to the computing phase of the CPU.
Let's say it can handle 4 TB with some difficulty...
|
|
|
|
|
|
