Update and News: www.burstcoin.de
>> New Feature German Poolmining Guide now available.
------
Help me to keep up the page and enhance the information for the german community. 220k to go in the Crowfund. Please do some donations!!!!
[/quote]
Topic: [ANN][BURST] Burst | Efficient HDD Mining | New 1.2.3 Fork block 92000 - page 364. (Read 2170648 times)
March 10, 2015, 02:43:56 AM
March 10, 2015, 01:51:39 AM
I agree with CIYAM.. we should change that "-" to something that can be copied on double-click. Decided to play with the code and figured out how to switch it over to "_" which happens to be copyable. Here try double clicking what will become my new wallet address: BURST_SYYS_QMGR_BCYN_BVQ2A
Except underscores don't select the whole thing when double clicking either.
Yup - I would recommend something like this:
BURSTxSYYSxQMGRxBCYNxBVQ2A
(doesn't look beautiful but at least works everywhere for selection)
March 10, 2015, 12:15:58 AM
I agree with CIYAM.. we should change that "-" to something that can be copied on double-click. Decided to play with the code and figured out how to switch it over to "_" which happens to be copyable. Here try double clicking what will become my new wallet address: BURST_SYYS_QMGR_BCYN_BVQ2A
Except underscores don't select the whole thing when double clicking either.
March 09, 2015, 11:46:08 PM
Whatever it is, but I think some big players must work out a way to mine big and dump burst price at poloniex these few days~~ 
You think people are going to start dumping burst? And therefore you'll be able to buy in even cheaper? Why the frowny face?
March 09, 2015, 11:26:50 PM
Any developers who've built the source code in windows? Finding it tricky.. the win-compile.sh is simply exactly the compile.sh, which is doesn't work because it's windows.. not linux and changing the name of the file doesn't magically make it work.
Anyone have experience with javac in bat files? Or Ant or Maven?
Thanks.
I agree with CIYAM.. we should change that "-" to something that can be copied on double-click. Decided to play with the code and figured out how to switch it over to "_" which happens to be copyable. Here try double clicking what will become my new wallet address: BURST_SYYS_QMGR_BCYN_BVQ2A
Everyone approve? We will of course have to give blockchain explorers a chance to make that same tweak.
Note for now, in my build either the "_" or the "-" is able to be input and verified as correct.. at least in theory once I can build the code.
Again, anyone able to help get the code to the point where we can run a .bat file and have the code be built.. maybe after another evening or two of playing with it I can get it working.
Anyone have experience with javac in bat files? Or Ant or Maven?
Thanks.
I agree with CIYAM.. we should change that "-" to something that can be copied on double-click. Decided to play with the code and figured out how to switch it over to "_" which happens to be copyable. Here try double clicking what will become my new wallet address: BURST_SYYS_QMGR_BCYN_BVQ2A
Everyone approve? We will of course have to give blockchain explorers a chance to make that same tweak.
Note for now, in my build either the "_" or the "-" is able to be input and verified as correct.. at least in theory once I can build the code.
Again, anyone able to help get the code to the point where we can run a .bat file and have the code be built.. maybe after another evening or two of playing with it I can get it working.
March 09, 2015, 10:06:18 PM
And someone just added a extra 5PB to the network? I don't think it's people just plotting out excess storage... ROI is up to two years now... I can only fathom someone found out a way to mine over cloud storage easily... or someone figured out how to mine without plotting.
I think it's some other factor that it causing these big fluctuations in the network. If a person did have access to a vast amount of data storage surely they would bring it online 10,50 or even 100TB at a time?
Just wanna say it must not be some odd thing. I, literally, just happened to get access to +2pb some time ago. Will not comment on how.
Of which you're gonna donate 500TB to ByteEnterprises' Cloud project! lol. jk.
Was thinking the whole thing, just you got the shipment settled.
PM incoming my brother!
March 09, 2015, 09:32:14 PM
from the specs he gave the plotting speed needs to be 800 MB/s to fill 16 TB / node in 6 hrs, which yields 8 TB mean per node.
Oh, plus mining concurrently with up to 70 MB/s for 1/4 of the time, ramping up from the beginning. Well, at least the stagger would be nice with ~60 GiB RAM. Oh, and I guess you need a handful of wallets to serve thousand miner instances, or did anyone notice funny numbers on the pools ?
I doubt you get a free trial with these compute- and io-capabilities, a thousand of those, repeatedly ? hmm..
Nonetheless, what a funky homework assignment !
i suggested him to run a small "wallet farm" and forgot he could simply limit the submitted deadlines to reduce the wallet compute load for verification.
the stagger should be no issue since this is ssd storage.
i never used the cloud myself but maybe he is able to rent the storage and can connect the 6h miners/plotters to it.
last time i have checked the cloud options it was'nt such expensive for large storage amounts simply because you pay for io cycles in combination with transfered volumes. if the stagger is high and on cloud plotted onto a connected permanent storage by the 6h machines with big staggers the read cycles should be reduced to only 8-32 io ops for 16tb each block depending on plotsizes.
i am not aware of their tos but if they do an audit (and i am sure they do for such loads) there will be a huge invoice for the compute nodes because the use is for sure against their tos.
but who knows....
hmm..per TB you need to scan 90 GB/day for mining, with a bandwith of ~4 MB/s if you want to finish <60s, duty cycle is 0.25 then.
If you scale this to a PB, you have 4 GB/s bandwith. Bytes, not Bits. Easy between racks, rather expensive between sites.
Just like the 90 TB traffic/day. Oops.
Mining at this scale needs to be close to the storage, preferably directly connected, obviously.
Whatever it is, but I think some big players must work out a way to mine big and dump burst price at poloniex these few days~~
March 09, 2015, 09:29:13 PM
And someone just added a extra 5PB to the network? I don't think it's people just plotting out excess storage... ROI is up to two years now... I can only fathom someone found out a way to mine over cloud storage easily... or someone figured out how to mine without plotting.
I think it's some other factor that it causing these big fluctuations in the network. If a person did have access to a vast amount of data storage surely they would bring it online 10,50 or even 100TB at a time?
Just wanna say it must not be some odd thing. I, literally, just happened to get access to +2pb some time ago. Will not comment on how.
Of which you're gonna donate 500TB to ByteEnterprises' Cloud project! lol. jk.
Was thinking the whole thing, just you got the shipment settled.
March 09, 2015, 09:17:06 PM
And someone just added a extra 5PB to the network? I don't think it's people just plotting out excess storage... ROI is up to two years now... I can only fathom someone found out a way to mine over cloud storage easily... or someone figured out how to mine without plotting.
I think it's some other factor that it causing these big fluctuations in the network. If a person did have access to a vast amount of data storage surely they would bring it online 10,50 or even 100TB at a time?
Just wanna say it must not be some odd thing. I, literally, just happened to get access to +2pb some time ago. Will not comment on how.
Of which you're gonna donate 500TB to ByteEnterprises' Cloud project! lol. jk.
March 09, 2015, 09:10:34 PM
And someone just added a extra 5PB to the network? I don't think it's people just plotting out excess storage... ROI is up to two years now... I can only fathom someone found out a way to mine over cloud storage easily... or someone figured out how to mine without plotting.
I think it's some other factor that it causing these big fluctuations in the network. If a person did have access to a vast amount of data storage surely they would bring it online 10,50 or even 100TB at a time?
Just wanna say it must not be some odd thing. I, literally, just happened to get access to +2pb some time ago. Will not comment on how.
March 09, 2015, 08:56:47 PM
Superminer enough to give you a BURST Hardon heh heh
https://www.backblaze.com/blog/backblaze-storage-pod-4/
Complete with parts lists and how to build it. Better have a lot of free cash though
Here was the old version...
https://www.youtube.com/watch?v=zqfkN7x-AfA
https://www.backblaze.com/blog/backblaze-storage-pod-4/
Complete with parts lists and how to build it. Better have a lot of free cash though
Here was the old version...
https://www.youtube.com/watch?v=zqfkN7x-AfA
March 09, 2015, 07:46:42 PM
The various sites quoting network size use very small block-ranges for averaging, not very helpful IMO due to their spikey-ness.
Good point. My calculator now uses 360 block average:
https://bchain.info/BURST/tools/calculator
March 09, 2015, 05:56:54 PM
from the specs he gave the plotting speed needs to be 800 MB/s to fill 16 TB / node in 6 hrs, which yields 8 TB mean per node.
Oh, plus mining concurrently with up to 70 MB/s for 1/4 of the time, ramping up from the beginning. Well, at least the stagger would be nice with ~60 GiB RAM. Oh, and I guess you need a handful of wallets to serve thousand miner instances, or did anyone notice funny numbers on the pools ?
I doubt you get a free trial with these compute- and io-capabilities, a thousand of those, repeatedly ? hmm..
Nonetheless, what a funky homework assignment !
i suggested him to run a small "wallet farm" and forgot he could simply limit the submitted deadlines to reduce the wallet compute load for verification.
the stagger should be no issue since this is ssd storage.
i never used the cloud myself but maybe he is able to rent the storage and can connect the 6h miners/plotters to it.
last time i have checked the cloud options it was'nt such expensive for large storage amounts simply because you pay for io cycles in combination with transfered volumes. if the stagger is high and on cloud plotted onto a connected permanent storage by the 6h machines with big staggers the read cycles should be reduced to only 8-32 io ops for 16tb each block depending on plotsizes.
i am not aware of their tos but if they do an audit (and i am sure they do for such loads) there will be a huge invoice for the compute nodes because the use is for sure against their tos.
but who knows....
hmm..per TB you need to scan 90 GB/day for mining, with a bandwith of ~4 MB/s if you want to finish <60s, duty cycle is 0.25 then.
If you scale this to a PB, you have 4 GB/s bandwith. Bytes, not Bits. Easy between racks, rather expensive between sites.
Just like the 90 TB traffic/day. Oops.
Mining at this scale needs to be close to the storage, preferably directly connected, obviously.
March 09, 2015, 05:45:47 PM
And someone just added a extra 5PB to the network? I don't think it's people just plotting out excess storage... ROI is up to two years now... I can only fathom someone found out a way to mine over cloud storage easily... or someone figured out how to mine without plotting.
I think it's some other factor that it causing these big fluctuations in the network. If a person did have access to a vast amount of data storage surely they would bring it online 10,50 or even 100TB at a time?
from the specs he gave the plotting speed needs to be 800 MB/s to fill 16 TB / node in 6 hrs, which yields 8 TB mean per node.
Oh, plus mining concurrently with up to 70 MB/s for 1/4 of the time, ramping up from the beginning. Well, at least the stagger would be nice with ~60 GiB RAM. Oh, and I guess you need a handful of wallets to serve thousand miner instances, or did anyone notice funny numbers on the pools ?
I doubt you get a free trial with these compute- and io-capabilities, a thousand of those, repeatedly ? hmm..
Nonetheless, what a funky homework assignment !
i suggested him to run a small "wallet farm" and forgot he could simply limit the submitted deadlines to reduce the wallet compute load for verification.
the stagger should be no issue since this is ssd storage.
i never used the cloud myself but maybe he is able to rent the storage and can connect the 6h miners/plotters to it.
last time i have checked the cloud options it was'nt such expensive for large storage amounts simply because you pay for io cycles in combination with transfered volumes. if the stagger is high and on cloud plotted onto a connected permanent storage by the 6h machines with big staggers the read cycles should be reduced to only 8-32 io ops for 16tb each block depending on plotsizes.
i am not aware of their tos but if they do an audit (and i am sure they do for such loads) there will be a huge invoice for the compute nodes because the use is for sure against their tos.
but who knows....
March 09, 2015, 05:42:18 PM
And someone just added a extra 5PB to the network? I don't think it's people just plotting out excess storage... ROI is up to two years now... I can only fathom someone found out a way to mine over cloud storage easily... or someone figured out how to mine without plotting.
I think it's some other factor that it causing these big fluctuations in the network. If a person did have access to a vast amount of data storage surely they would bring it online 10,50 or even 100TB at a time?
from the specs he gave the plotting speed needs to be 800 MB/s to fill 16 TB / node in 6 hrs, which yields 8 TB mean per node.
Oh, plus mining concurrently with up to 70 MB/s for 1/4 of the time, ramping up from the beginning. Well, at least the stagger would be nice with ~60 GiB RAM. Oh, and I guess you need a handful of wallets to serve thousand miner instances, or did anyone notice funny numbers on the pools ?
I doubt you get a free trial with these compute- and io-capabilities, a thousand of those, repeatedly ? hmm..
Nonetheless, what a funky homework assignment !
nicely summed up vaxman.
i suggested him to run a small "wallet farm" and forgot he could simply limit the submitted deadlines to reduce wallet compute load for verification.
the stagger should be no issue since this is ssd storage.
i never used the cloud myself but maybe he is able to rent the storage and can connect the miners to it.
last time i have checked the options it was'nt such expensive for large storage amounts simply because you pay for io cycles in combination with transfered volumes. if the stagger is high and on cloud plotted onto a connected permanent storage by the 6h machines with big staggers the read cycles should be reduced to only 8-32 for each block depending on plotsizes.
i am not aware of their tos but if they do an audit (and i am sure they do for such load) there will be a huge invoice for the compute nodes because the use is for sure against their tos.
but who knows....
Wow, I didn't have the info about the type of storage.
Which makes this even more improbable..
Unlimited CPU and IO is a no brainer, anyone can do the math. As a provider you somehow need to get your capex back in <2y or you're broke fast.
Add the opex for running 1k 16c/64GB boxes w/16TB SDD under full load for a single user and I bet he gets barely one test-cycle before controlling invalidates the free-ride tickets.
March 09, 2015, 05:32:14 PM
And someone just added a extra 5PB to the network? I don't think it's people just plotting out excess storage... ROI is up to two years now... I can only fathom someone found out a way to mine over cloud storage easily... or someone figured out how to mine without plotting.
I think it's some other factor that it causing these big fluctuations in the network. If a person did have access to a vast amount of data storage surely they would bring it online 10,50 or even 100TB at a time?
from the specs he gave the plotting speed needs to be 800 MB/s to fill 16 TB / node in 6 hrs, which yields 8 TB mean per node.
Oh, plus mining concurrently with up to 70 MB/s for 1/4 of the time, ramping up from the beginning. Well, at least the stagger would be nice with ~60 GiB RAM. Oh, and I guess you need a handful of wallets to serve thousand miner instances, or did anyone notice funny numbers on the pools ?
I doubt you get a free trial with these compute- and io-capabilities, a thousand of those, repeatedly ? hmm..
Nonetheless, what a funky homework assignment !
nicely summed up vaxman.
i suggested him to run a small "wallet farm" and forgot he could simply limit the submitted deadlines to reduce wallet compute load for verification.
the stagger should be no issue since this is ssd storage.
i never used the cloud myself but maybe he is able to rent the storage and can connect the miners to it.
last time i have checked the options it was'nt such expensive for large storage amounts simply because you pay for io cycles in combination with transfered volumes. if the stagger is high and on cloud plotted onto a connected permanent storage by the 6h machines with big staggers the read cycles should be reduced to only 8-32 for each block depending on plotsizes.
i am not aware of their tos but if they do an audit (and i am sure they do for such load) there will be a huge invoice for the compute nodes because the use is for sure against their tos.
but who knows....
March 09, 2015, 05:07:37 PM
And someone just added a extra 5PB to the network? I don't think it's people just plotting out excess storage... ROI is up to two years now... I can only fathom someone found out a way to mine over cloud storage easily... or someone figured out how to mine without plotting.
I don't think so, from my self-inflicted logalizer these are the network sizes for the last week, 360-block (1d) averaged, in TB:
15240.4
12857
14399.8
15732.5
12798.9
13660.2
15198.9
The various sites quoting network size use very small block-ranges for averaging, not very helpful IMO due to their spikey-ness.
March 09, 2015, 04:45:10 PM
Quote
give me
Quote
{"c":{"price":"0.00346021","market_cap":"0"},"ccex":{"price":"158","volume":"4.94184035","volume_burst":"2762333.3817643","market_cap":"0"},"trex":{"price":"152","volume":"0.82386345","volume_burst":"531105.70492874","market_cap":"0"},"polo":{"price":"159","volume":"2.91756645","volume_burst":"531105.70492874","market_cap":"0"},"price":"156","volume":"8.68327025","volume_burst":"5147663.170563","market_cap":"0"}
Ok it gave me good results hours ago, but now it seems it it broken (market cap is 0 and burst volume of trex and polo are equals !). Also is it possible to provide small but better API ?
Maybe a API to get market cap in currency you want, an other to extract price market you want or an average depending of argument you send, etc.
thanks
Edit, I'll write my own, and share it.
Edit 2 just learned Json_decode function in php... Things get much easier !
March 09, 2015, 04:41:54 PM
And someone just added a extra 5PB to the network? I don't think it's people just plotting out excess storage... ROI is up to two years now... I can only fathom someone found out a way to mine over cloud storage easily... or someone figured out how to mine without plotting.
I think it's some other factor that it causing these big fluctuations in the network. If a person did have access to a vast amount of data storage surely they would bring it online 10,50 or even 100TB at a time?
from the specs he gave the plotting speed needs to be 800 MB/s to fill 16 TB / node in 6 hrs, which yields 8 TB mean per node.
Oh, plus mining concurrently with up to 70 MB/s for 1/4 of the time, ramping up from the beginning. Well, at least the stagger would be nice with ~60 GiB RAM. Oh, and I guess you need a handful of wallets to serve thousand miner instances, or did anyone notice funny numbers on the pools ?
I doubt you get a free trial with these compute- and io-capabilities, a thousand of those, repeatedly ? hmm..
Nonetheless, what a funky homework assignment !
March 09, 2015, 02:34:42 PM
And someone just added a extra 5PB to the network? I don't think it's people just plotting out excess storage... ROI is up to two years now... I can only fathom someone found out a way to mine over cloud storage easily... or someone figured out how to mine without plotting.
I think it's some other factor that it causing these big fluctuations in the network. If a person did have access to a vast amount of data storage surely they would bring it online 10,50 or even 100TB at a time?
Jump to: