Did your client (bitminter) will work with this hardware-http://www.btcfpga.com/index.php?route=product/product&path=59&product_id=50 I'm on the way to buy that and i want to know is it possible to mine in your pool with bitminter software,becouse i like it. I know that you have support BFL Singles,but what's happened with another FPGA-s like this one and ZTEX-1.15y
Is there any info on how to control this board? BFL units use a virtual serial port over the USB connection. I'm going to add Icarus/Cairnsmore next. They also use serial port, so should be easy to add now that I have the serial port mining running with BFL. I'd like to add others too, I just want to add support for more serial port based devices first as it should be quick to do.
|
|
|
My blocks are still being merged, once they are done I just run it with my wallet info correct?
You mean your wallet program is still downloading blocks? You can send coins to the wallet before it finishes downloading all the blocks. You just can't see those coins in your wallet balance before it downloads the block that contains the transaction sending you coins.
|
|
|
Attack seems to have stopped. Let's hope it stays that way.
|
|
|
Should you find out who's behind it and submit proof I can possibly be able to return in kind, looks like I have a dosen or so wlans to abuse in addition to my own bandwith... and 7 computers to throw at it so an estimated 100-120Mb/s.
Not easy to find the identity of the attacker. The IRC and web post source (see below) is probably a hacked computer. Same with all the attacking computers. Message from attacker through the web contact form: Sent from IP: 109.123.96.23
SEND US 100 USD ON LIBETYRESERVE ACCOUNT U3892021
AND WE WILL STOP THE ATTACK.
IF YOU DONT PAY US WE WILL KEEP ATTACKING FOR ONE WEEK OR 2 LETTING YOUR BUSINESS DIED
Well, I will lose the 2 USD I could have made in those 2 weeks. The attackers will lose whatever a 2 week DDoS attack of this magnitude costs. I think that's more than 2 USD At least there's that.
|
|
|
Bit of a DDOS attack going on now... From IRC: 14:18 -!- DDOSATTACK [6d7b6017@gateway/web/freenode/ip.109.123.96.23] has joined #bitminter 14:19 -!- DDOSATTACK [6d7b6017@gateway/web/freenode/ip.109.123.96.23] has quit [Client Quit]
Started 7-8 minutes after that. About 90% packet loss to the server at the moment.
|
|
|
When I start the speed rises up to its supposed maximum. Temperature is between 45° and 55° C. After a short while it seems to start throttling and these messages show up.
Yeah, if it is throttling it's not responding to commands for some seconds. That's why you are getting the timeouts. It will still run although the messages can be annoying. But it may be best to avoid throttling. Perhaps improve cooling, or use a firmware with lower hash rate.
|
|
|
I am gonna throw 20 singles at ur pool soon..
Awesome i want different people to have acess to my account..so how can i have an "normal" login?
OInly openID are avaliable..i dont want the hassel making fake accounts for that .
Only OpenID logins available for now. You could create an OpenID account somewhere and give them that password. But do you really want a whole bunch of people to have access to your account? Do you want all of them to have full access, including the ability to transfer all coins in the account to their wallet? Is the site down or is it just me?
I just restarted the web app - I don't think it was down many seconds? I try not to restart often, I know it's annoying to have to log in again. But I needed to clear out old database connections to do some cleanup in the database. Such a relief to see transactions that took several seconds become instant.
|
|
|
Version 1.2.0 is out. Thanks to everyone who helped test it, especially Fefox and Phraust. This took a lot more effort than I had anticipated, but now it's done and we have a much improved miner. Changes from 1.1.2 to 1.2.0: - Support FPGA devices from Butterfly Labs (BFL)
- Network subsystem rewritten: fast work download over multiple pipelined
connections to be able to handle higher hash rates. Also yields fewer rejected proofs of work. - Added minimize/maximize buttons on each device. The compact display
is handy if you have a long list of devices. Remembered across restarts. - The start button of the "total" device now starts only the visible
devices instead of all devices the first time you click it. The next time you use it to start devices it will resume those you stopped with it, like it always did. The start button on the status bar works the same. - Adjusted expected hash rate (green zone) for Juniper GPU
- Fixed some minor GUI bugs
Try it out: Windows 8 note:Version 1.2.0 will not run on current Windows 8 beta versions with current Java releases because of a problem with asynchronous I/O running Java on Windows 8. Until this is fixed in either Java or Windows, if you want to run the miner on Windows 8 you will have to use the previous release: http://bitminter.com/client/1.1.2/bitminter.jnlp
|
|
|
Beta 12 is out, with 2 changes: - Fixed memory leak that would lead to high memory and CPU consumption over time, faster if you had high hash rate
- Made the network code more likely to go from 1 to 2 parallel connections
Testing this for 3-4 hours shows very low CPU usage with BFL singles and only about 0.05% rejected proofs of work. (CPU use will be a bit higher on Mac OS X than Windows and Linux because I had to implement a dirty workaround to avoid a bug in the serial port library) If anyone has a chance to try this on the Raspberry Pi or other ARM-based devices, please let me know. In theory it may work, but I have a feeling the miner will be a tight fit for the low amount of memory on these devices. There's almost certainly a workaround. PM me if you want me to look at the relevant code.
According to Oracle's bug tracker there's no workaround. I guess the workaround for us is having synchronous networking for running on Windows 8 until they fix this issue. I could offer older versions (suggested by btc196) or offer a synchronous/asynchronous switch in config (suggested by LordMord). Making older versions available is the quickest solution, and something I was planning on doing at some point anyway. But it does mean new releases won't work on Windows 8 until Microsoft or Oracle implement a fix. Anyone planning on running FPGAs on Windows 8 beta?
|
|
|
Running Windows 8. But that was no problem for beta 9.
Beta 9 was using synchronous network I/O. Beta 11 has asynchronous I/O. Apparently when doing an asynchronous connect Java will call getsockname() before the connection is established. This works on every operating system except Windows 8. Are you using the latest, Java 7, update 4? This could become an annoying issue. There's no workaround either.
|
|
|
One problem found in beta 11. On some systems (high hash rate?) after running for some time the CPU usage goes up and it starts running out of work. Investigating this now. Question - would that new version help stale proof of work at all? I have 4 BFL's running, and don't ever get a lack of work, but do get stale proof of work from time to time.
Yeah, it can use multiple connections and it gives priority to uploading proofs of work over downloading more work, so it might reduce rejects a bit. I get this error with beta 11: Connection failed: Invalid argument: getsockname
Do you have a firewall blocking the program? Are you running on Windows 8?
|
|
|
I just uploaded beta 11. It fixes a thread scheduling bug in beta 10 that could cause devices to run out of work when there was still work in the queue. Beta 11 seems to run fine on 6 BFL singles, 5.2 GH/s, with low stales. Looking good so far.
|
|
|
I.m testing your v1.2.0 beta10 with one Bit Force and I get: batch acquisition of 0 triggers.
What does it mean?
It's just the scheduling subsystem (see the automation part of the options) that is a bit chatty. It's nothing to worry about and you should only see it if you have verbose logging turned on.
|
|
|
v1.2.0 beta10 is out. This version can download work much faster, using multiple pipelined connections to the server. This should prevent running out of work when you have a high hash rate on the computer, like 9-10 BFL singles or more. All the network code has been rewritten completely. I've been running this a few days myself, but it needs some more testing before finalizing v1.2.0. Please check it out, especially interesting if you can try it on high latency connections (maybe big file transfers running that slow down your internet connection) or an unstable network where the connection comes and goes. Let me know how it runs.
|
|
|
From Slipbye (who is locked in the newbie section):
If you plug in your monitor into the gfx card thats only hashing at 50% then in CCC change the core clock up or down a couple mhz, just anything to make a change, hit Apply it should fix it. Do the same with the last Card again.
|
|
|
Whats with block 183419 "minted by external" ?
Long running transactions in the database kept the pool from recording who created the block. Lacking this info, the pool is acting like the block was created outside the pool. No worries, the block is fine, all proofs of work recorded and coins are split out to miners as usual. Sorry for the confusion. I'm looking into why some transactions are running slowly. May have to tune the database a bit. Good thing this didn't happen during the mint race.
|
|
|
I missed the renewal time for the SSL certificate today. It's annoying they have a 48 hour window to renew. Now fixed, and https works without warnings. Sorry for the inconvenience. Looks very nice, thanks!
Glad you like it
|
|
|
|