Topic: [ANN][BURST] Burst | Efficient HDD Mining | New 1.2.3 Fork block 92000 - page 663. (Read 2171078 times)
October 07, 2014, 08:56:13 AM
October 07, 2014, 08:09:56 AM
Yellow: it seems that reading plots is multithreaded even on one drive
Red: I have no ideas
For the Red, I notified this, some post before this, but no replies..
October 07, 2014, 07:41:16 AM
The java miner reads sequentially. Uray's miner uses more threads to read files. If you have a single file it will have only 1 thread. But if you have 2, it will create a second thread for that file.
If it's true, the performance will be very bad unless there's a single big file on each disk (I haven't used Uray's miner).
dcct's miner creates one thread per directory. If you create 1 directory per 1 hdd, they will be read in parallel without interfering with each other.
October 07, 2014, 07:34:32 AM
Yellow: it seems that reading plots is multithreaded even on one drive
Red: I have no ideas
October 07, 2014, 06:45:40 AM
Great news , thanks @unso !
October 07, 2014, 06:37:11 AM
October 07, 2014, 06:25:18 AM
The java miner consumes the lowest electricity out of all of them, but has memory hogging/crash issues.
Java miner is a "RAM miner" whereas those new miner that promise low RAM usage is a "CPU miner". Running 100% CPU consume more electricity.
I am still using java miner because i have low end CPU on my rig and it perform badly on the new miner. Java finish reading plot and new miner still reading plot and take quite a while to finish reading the whole plot because CPU is the bottleneck. More CPU power faster reading. If u are running hundreds of TB, an i7 would probably be needed for new miner.
Anyway , i found a way to solve the RAM issue and my rigs has been running smoothly without any manual input from me for a week and counting. Its pretty much auto pilot right now using java miner and payout from pool is very stable because of zero downtime.
I'm running a Core i5(mining 14.5TB) and an old AMD Phenom II x4(mining 10.5TB), both run about 28%~40% usage on average. Just a little bit more power than the Java miner, I was running the script to do auto-restart with the java. Still the java miner would stop working at times and both setup have 8GB of RAM.
October 07, 2014, 06:16:59 AM
Good job fenape!!!
let's go to help on burst!!!
let's go to help on burst!!!
Thanks for the rt
October 07, 2014, 06:15:27 AM
Nevril, thank you for translation My english is poor. Looks like there`s some problems because of it.
Now i`m preparing fully staggered and defragmented plotfiles. Will see...
My english is poor. Looks like there`s some problems because of it....
Uray's miner uses more threads to read files. If you have a single file it will have only 1 thread. But if you have 2, it will create a second thread for that file.
...
Are you shure? Can you show me dev`s post about it or piece of code?Uray's miner uses more threads to read files. If you have a single file it will have only 1 thread. But if you have 2, it will create a second thread for that file.
...
Now i`m preparing fully staggered and defragmented plotfiles. Will see...
Miners dev name is Uray. https://github.com/uraymeiviar/burst-miner/releases
You can find the source there.
I am not sure if Uray specified on the forum that the miners uses multiple threads or in a private message from our conversations.
October 07, 2014, 06:08:57 AM
First of all I want to specify that I completely agree with fanepatent.
You cannot drop in here, newbie or not, and insult. It isn't neither a matter of who you're talking too. It could have been one of the most active and helpful people like fane is since the beginning or a completely unknown guy (like me).
Said so I'll try to reformulate your question because I think it is an interesting one, just because of this.
I think Tex is asking if, when having two different plots in the same HDD/directory, uray's miner reads them concurrently instead of sequentially.
If this is the case, no matter the stagger size, the HDD will be put under a lot of pressure with jumps back and forth.
To reply, partially, to Tex (I don't use uray's so I don't know). The Dev's miner is the Java one (POC Miner), and it reads sequentially for sure. You may want to have a try with it on Dev's pool or solo and see if you have the same issue.
You cannot drop in here, newbie or not, and insult. It isn't neither a matter of who you're talking too. It could have been one of the most active and helpful people like fane is since the beginning or a completely unknown guy (like me).
Said so I'll try to reformulate your question because I think it is an interesting one, just because of this.
I think Tex is asking if, when having two different plots in the same HDD/directory, uray's miner reads them concurrently instead of sequentially.
If this is the case, no matter the stagger size, the HDD will be put under a lot of pressure with jumps back and forth.
To reply, partially, to Tex (I don't use uray's so I don't know). The Dev's miner is the Java one (POC Miner), and it reads sequentially for sure. You may want to have a try with it on Dev's pool or solo and see if you have the same issue.
Yes I got that and it can be seen in my "good answer" after being insulted...
The java miner reads sequentially. Uray's miner uses more threads to read files. If you have a single file it will have only 1 thread. But if you have 2, it will create a second thread for that file.
The files are split into segments. Each file having x amount of segments and each segment represents the one stagger.
Your 1 file has 256 segments. Yes, it reads that well. But if you make another one your HDD will have another 256 seeks to do at the same time.
Your 1 file has 256 segments. Yes, it reads that well. But if you make another one your HDD will have another 256 seeks to do at the same time.
It seems that Blago's Windows miner is more refined and spreads the work evenly across all threads, regardless of the number of plots.
I haven't used Blago's miner. I need to do a fresh install. So I can't talk about that.
i offer skype support to miners with problem on setting up theri rigs via skype from 09.00AM[GMT+2] to 06:00PM[GMT+2]
just search HelpingBurstMiners over Skype!!!
hope this could help Burst community!
just search HelpingBurstMiners over Skype!!!
hope this could help Burst community!
Skype support also available from me too. Well, mostly for the pool, but I'm sure people will start using it for general questions.
Skype link can be found on pool page under How to and How it works links
Good job fenape!!!
let's go to help on burst!!!
October 07, 2014, 05:54:45 AM
The java miner consumes the lowest electricity out of all of them, but has memory hogging/crash issues.
Java miner is a "RAM miner" whereas those new miner that promise low RAM usage is a "CPU miner". Running 100% CPU consume more electricity.
I am still using java miner because i have low end CPU on my rig and it perform badly on the new miner. Java finish reading plot and new miner still reading plot and take quite a while to finish reading the whole plot because CPU is the bottleneck. More CPU power faster reading. If u are running hundreds of TB, an i7 would probably be needed for new miner.
Anyway , i found a way to solve the RAM issue and my rigs has been running smoothly without any manual input from me for a week and counting. Its pretty much auto pilot right now using java miner and payout from pool is very stable because of zero downtime.
October 07, 2014, 05:41:01 AM
First of all I want to specify that I completely agree with fanepatent.
You cannot drop in here, newbie or not, and insult. It isn't neither a matter of who you're talking too. It could have been one of the most active and helpful people like fane is since the beginning or a completely unknown guy (like me).
Said so I'll try to reformulate your question because I think it is an interesting one, just because of this.
I think Tex is asking if, when having two different plots in the same HDD/directory, uray's miner reads them concurrently instead of sequentially.
If this is the case, no matter the stagger size, the HDD will be put under a lot of pressure with jumps back and forth.
To reply, partially, to Tex (I don't use uray's so I don't know). The Dev's miner is the Java one (POC Miner), and it reads sequentially for sure. You may want to have a try with it on Dev's pool or solo and see if you have the same issue.
You cannot drop in here, newbie or not, and insult. It isn't neither a matter of who you're talking too. It could have been one of the most active and helpful people like fane is since the beginning or a completely unknown guy (like me).
Said so I'll try to reformulate your question because I think it is an interesting one, just because of this.
I think Tex is asking if, when having two different plots in the same HDD/directory, uray's miner reads them concurrently instead of sequentially.
If this is the case, no matter the stagger size, the HDD will be put under a lot of pressure with jumps back and forth.
To reply, partially, to Tex (I don't use uray's so I don't know). The Dev's miner is the Java one (POC Miner), and it reads sequentially for sure. You may want to have a try with it on Dev's pool or solo and see if you have the same issue.
Yes I got that and it can be seen in my "good answer" after being insulted...
The java miner reads sequentially. Uray's miner uses more threads to read files. If you have a single file it will have only 1 thread. But if you have 2, it will create a second thread for that file.
The files are split into segments. Each file having x amount of segments and each segment represents the one stagger.
Your 1 file has 256 segments. Yes, it reads that well. But if you make another one your HDD will have another 256 seeks to do at the same time.
Your 1 file has 256 segments. Yes, it reads that well. But if you make another one your HDD will have another 256 seeks to do at the same time.
It seems that Blago's Windows miner is more refined and spreads the work evenly across all threads, regardless of the number of plots.
I haven't used Blago's miner. I need to do a fresh install. So I can't talk about that.
i offer skype support to miners with problem on setting up theri rigs via skype from 09.00AM[GMT+2] to 06:00PM[GMT+2]
just search HelpingBurstMiners over Skype!!!
hope this could help Burst community!
just search HelpingBurstMiners over Skype!!!
hope this could help Burst community!
Skype support also available from me too. Well, mostly for the pool, but I'm sure people will start using it for general questions.
Skype link can be found on pool page under How to and How it works links
After using the java miner, Uray's miner, dcct's miner and Blago's miner: Blago worked best for me on Windows and dcct is best on Linux. The java miner consumes the lowest electricity out of all of them, but has memory hogging/crash issues.
October 07, 2014, 05:35:15 AM
Nevril, thank you for translation My english is poor. Looks like there`s some problems because of it.
Now i`m preparing fully staggered and defragmented plotfiles. Will see...
My english is poor. Looks like there`s some problems because of it....
Uray's miner uses more threads to read files. If you have a single file it will have only 1 thread. But if you have 2, it will create a second thread for that file.
...
Are you shure? Can you show me dev`s post about it or piece of code?Uray's miner uses more threads to read files. If you have a single file it will have only 1 thread. But if you have 2, it will create a second thread for that file.
...
Now i`m preparing fully staggered and defragmented plotfiles. Will see...
October 07, 2014, 05:28:41 AM
First of all I want to specify that I completely agree with fanepatent.
You cannot drop in here, newbie or not, and insult. It isn't neither a matter of who you're talking too. It could have been one of the most active and helpful people like fane is since the beginning or a completely unknown guy (like me).
Said so I'll try to reformulate your question because I think it is an interesting one, just because of this.
I think Tex is asking if, when having two different plots in the same HDD/directory, uray's miner reads them concurrently instead of sequentially.
If this is the case, no matter the stagger size, the HDD will be put under a lot of pressure with jumps back and forth.
To reply, partially, to Tex (I don't use uray's so I don't know). The Dev's miner is the Java one (POC Miner), and it reads sequentially for sure. You may want to have a try with it on Dev's pool or solo and see if you have the same issue.
You cannot drop in here, newbie or not, and insult. It isn't neither a matter of who you're talking too. It could have been one of the most active and helpful people like fane is since the beginning or a completely unknown guy (like me).
Said so I'll try to reformulate your question because I think it is an interesting one, just because of this.
I think Tex is asking if, when having two different plots in the same HDD/directory, uray's miner reads them concurrently instead of sequentially.
If this is the case, no matter the stagger size, the HDD will be put under a lot of pressure with jumps back and forth.
To reply, partially, to Tex (I don't use uray's so I don't know). The Dev's miner is the Java one (POC Miner), and it reads sequentially for sure. You may want to have a try with it on Dev's pool or solo and see if you have the same issue.
Yes I got that and it can be seen in my "good answer" after being insulted...
The java miner reads sequentially. Uray's miner uses more threads to read files. If you have a single file it will have only 1 thread. But if you have 2, it will create a second thread for that file.
The files are split into segments. Each file having x amount of segments and each segment represents the one stagger.
Your 1 file has 256 segments. Yes, it reads that well. But if you make another one your HDD will have another 256 seeks to do at the same time.
Your 1 file has 256 segments. Yes, it reads that well. But if you make another one your HDD will have another 256 seeks to do at the same time.
It seems that Blago's Windows miner is more refined and spreads the work evenly across all threads, regardless of the number of plots.
I haven't used Blago's miner. I need to do a fresh install. So I can't talk about that.
i offer skype support to miners with problem on setting up theri rigs via skype from 09.00AM[GMT+2] to 06:00PM[GMT+2]
just search HelpingBurstMiners over Skype!!!
hope this could help Burst community!
just search HelpingBurstMiners over Skype!!!
hope this could help Burst community!
Skype support also available from me too. Well, mostly for the pool, but I'm sure people will start using it for general questions.
Skype link can be found on pool page under How to and How it works links
October 07, 2014, 05:25:29 AM
i offer skype support to miners with problem on setting up theri rigs via skype from 09.00AM[GMT+2] to 06:00PM[GMT+2]
just search HelpingBurstMiners over Skype!!!
if you do not find it just search UnSoInDoVo, always on skype.
hope this could help Burst community!
just search HelpingBurstMiners over Skype!!!
if you do not find it just search UnSoInDoVo, always on skype.
hope this could help Burst community!
October 07, 2014, 05:25:04 AM
First of all I want to specify that I completely agree with fanepatent.
You cannot drop in here, newbie or not, and insult. It isn't neither a matter of who you're talking too. It could have been one of the most active and helpful people like fane is since the beginning or a completely unknown guy (like me).
Said so I'll try to reformulate your question because I think it is an interesting one, just because of this.
I think Tex is asking if, when having two different plots in the same HDD/directory, uray's miner reads them concurrently instead of sequentially.
If this is the case, no matter the stagger size, the HDD will be put under a lot of pressure with jumps back and forth.
To reply, partially, to Tex (I don't use uray's so I don't know). The Dev's miner is the Java one (POC Miner), and it reads sequentially for sure. You may want to have a try with it on Dev's pool or solo and see if you have the same issue.
You cannot drop in here, newbie or not, and insult. It isn't neither a matter of who you're talking too. It could have been one of the most active and helpful people like fane is since the beginning or a completely unknown guy (like me).
Said so I'll try to reformulate your question because I think it is an interesting one, just because of this.
I think Tex is asking if, when having two different plots in the same HDD/directory, uray's miner reads them concurrently instead of sequentially.
If this is the case, no matter the stagger size, the HDD will be put under a lot of pressure with jumps back and forth.
To reply, partially, to Tex (I don't use uray's so I don't know). The Dev's miner is the Java one (POC Miner), and it reads sequentially for sure. You may want to have a try with it on Dev's pool or solo and see if you have the same issue.
Yes I got that and it can be seen in my "good answer" after being insulted...
The java miner reads sequentially. Uray's miner uses more threads to read files. If you have a single file it will have only 1 thread. But if you have 2, it will create a second thread for that file.
The files are split into segments. Each file having x amount of segments and each segment represents the one stagger.
Your 1 file has 256 segments. Yes, it reads that well. But if you make another one your HDD will have another 256 seeks to do at the same time.
Your 1 file has 256 segments. Yes, it reads that well. But if you make another one your HDD will have another 256 seeks to do at the same time.
It seems that Blago's Windows miner is more refined and spreads the work evenly across all threads, regardless of the number of plots.
October 07, 2014, 05:24:29 AM
First of all I want to specify that I completely agree with fanepatent.
You cannot drop in here, newbie or not, and insult. It isn't neither a matter of who you're talking too. It could have been one of the most active and helpful people like fane is since the beginning or a completely unknown guy (like me).
Said so I'll try to reformulate your question because I think it is an interesting one, just because of this.
I think Tex is asking if, when having two different plots in the same HDD/directory, uray's miner reads them concurrently instead of sequentially.
If this is the case, no matter the stagger size, the HDD will be put under a lot of pressure with jumps back and forth.
To reply, partially, to Tex (I don't use uray's so I don't know). The Dev's miner is the Java one (POC Miner), and it reads sequentially for sure. You may want to have a try with it on Dev's pool or solo and see if you have the same issue.
You cannot drop in here, newbie or not, and insult. It isn't neither a matter of who you're talking too. It could have been one of the most active and helpful people like fane is since the beginning or a completely unknown guy (like me).
Said so I'll try to reformulate your question because I think it is an interesting one, just because of this.
I think Tex is asking if, when having two different plots in the same HDD/directory, uray's miner reads them concurrently instead of sequentially.
If this is the case, no matter the stagger size, the HDD will be put under a lot of pressure with jumps back and forth.
To reply, partially, to Tex (I don't use uray's so I don't know). The Dev's miner is the Java one (POC Miner), and it reads sequentially for sure. You may want to have a try with it on Dev's pool or solo and see if you have the same issue.
Yes I got that and it can be seen in my "good answer" after being insulted...
The java miner reads sequentially. Uray's miner uses more threads to read files. If you have a single file it will have only 1 thread. But if you have 2, it will create a second thread for that file.
The files are split into segments. Each file having x amount of segments and each segment represents the one stagger.
Your 1 file has 256 segments. Yes, it reads that well. But if you make another one your HDD will have another 256 seeks to do at the same time.
Your 1 file has 256 segments. Yes, it reads that well. But if you make another one your HDD will have another 256 seeks to do at the same time.
Think I missed your reply, ok... Tex has it's answer.
October 07, 2014, 05:21:07 AM
First of all I want to specify that I completely agree with fanepatent.
You cannot drop in here, newbie or not, and insult. It isn't neither a matter of who you're talking too. It could have been one of the most active and helpful people like fane is since the beginning or a completely unknown guy (like me).
Said so I'll try to reformulate your question because I think it is an interesting one, just because of this.
I think Tex is asking if, when having two different plots in the same HDD/directory, uray's miner reads them concurrently instead of sequentially.
If this is the case, no matter the stagger size, the HDD will be put under a lot of pressure with jumps back and forth.
To reply, partially, to Tex (I don't use uray's so I don't know). The Dev's miner is the Java one (POC Miner), and it reads sequentially for sure. You may want to have a try with it on Dev's pool or solo and see if you have the same issue.
You cannot drop in here, newbie or not, and insult. It isn't neither a matter of who you're talking too. It could have been one of the most active and helpful people like fane is since the beginning or a completely unknown guy (like me).
Said so I'll try to reformulate your question because I think it is an interesting one, just because of this.
I think Tex is asking if, when having two different plots in the same HDD/directory, uray's miner reads them concurrently instead of sequentially.
If this is the case, no matter the stagger size, the HDD will be put under a lot of pressure with jumps back and forth.
To reply, partially, to Tex (I don't use uray's so I don't know). The Dev's miner is the Java one (POC Miner), and it reads sequentially for sure. You may want to have a try with it on Dev's pool or solo and see if you have the same issue.
Yes I got that and it can be seen in my "good answer" after being insulted...
The java miner reads sequentially. Uray's miner uses more threads to read files. If you have a single file it will have only 1 thread. But if you have 2, it will create a second thread for that file.
The files are split into segments. Each file having x amount of segments and each segment represents the one stagger.
Your 1 file has 256 segments. Yes, it reads that well. But if you make another one your HDD will have another 256 seeks to do at the same time.
Your 1 file has 256 segments. Yes, it reads that well. But if you make another one your HDD will have another 256 seeks to do at the same time.
October 07, 2014, 05:18:44 AM
Downloaded uray`s miner (ver 1.1 R4).
One plotfile 1048576_4096 (256GB) - reading fast and smooth.
Two such files - HDD is going crazy. Reading slow and loud.
Is it normal for single folder/drive (single path in config)?
One plotfile 1048576_4096 (256GB) - reading fast and smooth.
Two such files - HDD is going crazy. Reading slow and loud.
Is it normal for single folder/drive (single path in config)?
With that stagger kinda is. It needs 256 seeks to read that file.
HDD was empty. Fragmentation = 0.
A single file with 256 fragments. Each fragment containing 4096 nonces.
Your drive needs to go through all those 256 fragments until it finds the right nonce.
Drop the aggressive and idiotic tone if you want help. I am with this project since it was launched. I think I know what I am talking about.
Now if you don't know how to ask for help and how to read the answers go back to shitcoins
The files are split into segments. Each file having x amount of segments and each segment represents the one stagger.
Your 1 file has 256 segments. Yes, it reads that well. But if you make another one your HDD will have another 256 seeks to do at the same time.
Loud and clear?
We do not tolerate your kind here. There are 660 pages of chatter and not one post had the tone you used.
I told you before. Most newbies are ignorant and ungrateful. They only care for themselves first. That's why I ignore most of similar questions from newbie in this thread. He doesn't even realize that his HDD might be dead soon, who cares?
First of all, your account. Second of all, thats a newbie question.
And now lets get back to where we were.
After you called me an idiot and I took the time to answer despite that, you just tell me that you will wait for the dev to come and answer?
Are you for real?
It`s not a newbie question. I`m asking not about stagger, seeking, fragmentation, hdd health, etc.
I didn`t call you an idiot. It was a question
Yes i`m 4 real. Question is about dev`s miner. Not about Burst basics.
The way you imply the question has a very rude context and tone, even if you meant not to be. If you don't like certain people's answer, then don't take offense and just find/wait for the answer that satisfies you. If you want to wait for the dev of the miner, then just be patient and wait until then.
October 07, 2014, 05:18:35 AM
Skype support on http://mine.burstcoin.info is now available.
If you have any questions or problems click the "Chat on Skype" link below the How to mine and How it works links.
Also email is available for support. But I'd rather not post it on a public forum.
Spread the news, bring miners, buy shares and see them grow in value.
If you have any questions or problems click the "Chat on Skype" link below the How to mine and How it works links.
Also email is available for support. But I'd rather not post it on a public forum.
Spread the news, bring miners, buy shares and see them grow in value.
just retweet the help skype channel to my followers
hope someone else can do the same!!!!
Jump to: