I hear some people have had issues with this, but I got it working in Linux with one restriction.
I set the ATI card as my primary adapter (ie: plugged the monitor in to it...). For whatever odd reason, you can't use OpenCL unless the card is up and initted. CUDA does not have this limitation.
So with the ATI card as primary and X started on it, I was able to use poclbm with the --platform flag to use both the ATI card and an NVIDIA card.
Since then, I haven't tried again.
|
|
|
Copypasta from another thread... Had I known this thread existed, I would've started here. #!/bin/bash while [ 1 ]; do echo "<html><head><meta http-equiv=\"refresh\" content=\"5\"/><body>$(aticonfig --adapter=all --odgt | sed ':a;N;$!ba;s/\n/<br\/>/g')<body></html>" | nc -l 9001 done
I realized we have a lot of new Linux users out there, so here's a simple script to keep you away from the GUI. Put the above in a file, chmod u+x it, run it. That's it. Connect with a browser to whatever the host is, port 9001. Requires netcat (nc). Feel free to replace "aticonfig --adapter=all --odgt" with whatever you like - use tee to redirect your miner output or whatever you want to a file, then push that through netcat. Let's say you've got a script called startminer.sh that starts poclbm (or whatever). You can use the following to start it: startminer.sh | tee -a minerlog.log Then use - #!/bin/bash while [ 1 ]; do echo "<html><head><meta http-equiv=\"refresh\" content=\"5\"/><body>$(sed ':a;N;$!ba;s/\n/<br\/>/g' < minerlog.log)<body></html>" | nc -l 9001 done
For ssh sessions, don't forget that happy watch command - watch -n 1 aticonfig --adapter=all --odgt Will continuously print to the screen the output from aticonfig every 1 second. And finally, if you're using eligius, don't forget there are both EU and US servers. I haven't been able to test the following script... eligius servers haven't gone down. But the idea is if you had a line in your /etc/hosts like: 173.242.112.67 eligius.mining And pointed a miner at http://eligius.mining:8337 while the script below ran..... if the hashrate for that server hit 0 or was otherwise unreachable it should "failover" to the other server. #!/bin/bash
EUHTTP="http://eligius.st/~luke-jr/raw/eu/hashrate.txt" USHTTP="http://eligius.st/~luke-jr/raw/us/hashrate.txt"
EUIP="85.25.78.8" USIP="173.242.112.67"
export EUIP export USIP
CURRENT=$USHTTP
echo $EUIP echo $USIP
while [ 1 == 1 ]; do echo "Checking..."
HR=$(wget -q -O - $CURRENT)
if [ -z "$HR" ] then HR="0" fi
echo "Hashrate $HR"
if [ $CURRENT == $USHTTP ] then if [ $HR == 0 ] then cp /etc/hosts /tmp/hosts sed s/$USIP/$EUIP/ < /tmp/hosts > /etc/hosts echo "Switched to Europe." CURRENT=$EUHTTP fi elif [ $CURRENT == $EUHTTP ] then if [ $HR == 0 ] then cp /etc/hosts /tmp/hosts sed s/$EUIP/$USIP/ < /tmp/hosts > /etc/hosts echo "Switched to US." CURRENT=$USHTTP fi fi sleep 5m done
|
|
|
Good point. Did not find that thread when I searched for "linux scripts" before.
|
|
|
All requests are over JSON/HTTP so only TCP.... You can see the status of the eligius servers and uptime at http://eligius.st/~artefact2/There was a recent bump in St. Eligius US.
|
|
|
Windows, eh?
Did you enable crossfire? You don't want that to mine...
|
|
|
Sounds like a great opportunity for you to provide one! I didn't ignore your original request - I've been digging around in aticonfig and pplib stuff but haven't found anything that gives me an amperage or estimated power use - only voltages.
|
|
|
I'm surprised no one mentioned the -q flag… Try using -q 4 or -q 5, and see if it still happens as frequently. Basically it tells your miner to use a larger work queue.
This. Bumped up the -q setting to 5 and that idle message disappeared. Good call, Artefact2.
|
|
|
Ya, wasn't really intending to insult his English. Just making random quotes. Anyway, point being, don't jump on the "every pool is DDoS OMG" bandwagon until someone confirms something. Or, even better, go solo! Supa, can you please explain the link between bitcoin & nodes and nodes & economic stability Did you want a dissertation or.... ? In a simplistic nutshell package, more nodes mean you have more transports for the peer-to-peer traffic of bitcoin. It was intended that these nodes be rewarded for performing work to keep the chain and transactions verified and correct without a single point of authority (or multiple points of trusted authority). More nodes means more data can be transported faster. More nodes actually performing verification (aka mining) translates to strength in the network. Pools provide the opposite. "Weekend warriors" can sweep into the bitcoin project with their insatiable requirement for instant gratification (that is, rewards from pool mining rather than long-term payoffs). These individuals can drastically alter price based on them having a bad day. For example, let's say 30 people that have a few hundred coins decide "eff bitcoin this is stupid I can't use deepbit DDoS Russian Chinese blaaaah!" and go piss away their coins on an exchange for nearly nothing. Deflation happens. If a 51% exploit is achieved believable... deflation will happen. If people start losing transactions.... deflation will happen. If downloading the block chain takes a year (current client requires the whole block chain) then no layman consumers will want to use it. Ditto for actually trying to find a node to submit a transaction to. Deflation will happen. More miners and less nodes reduce the stability and trust in the network. Now the opposite... Nodes are available and fast around the world. Transactions are nearly instant and there are hundreds of thousands of machines maintaining and verifying transactions. Blocks are easily downloaded and nodes are abundant. More nodes and less miners increase the stability and trust in the network (and by proxy - overall interest and willingness to use Bitcoin ~ stabilizing prices). If the network is stable, fast and trustworthy at all times.... Inflation can happen. Or at least stability.
|
|
|
There's this -
aticonfig --pplib-cmd "get activity" Current Activity is Core Clock: 1005MHZ Memory Clock: 225MHZ VDDC: 1088 Activity: 99 percent Performance Level: 2 Bus Speed: 2500 Bus Lanes: 16 Maximum Bus Lanes: 16
But there's no current amp or wattage. What Windows thing gives you a power consumption listing?
|
|
|
In my experience, if you use the -b thing, you can't adjust with aticonfig afterward. Which is great if you have 100% decided on clocks, but not so great if you're still fiddling. I just leave the GUI running.
|
|
|
Ya, wasn't really intending to insult his English. Just making random quotes. Anyway, point being, don't jump on the "every pool is DDoS OMG" bandwagon until someone confirms something. Or, even better, go solo!
|
|
|
"I is and or likes pie" - supa
|
|
|
AMDOverDriveCtrl *causing* freezing sounds like a red herring to me....
It's more likely you were setting invalid clocks or otherwise making your card unhappy.
Run AMDOverDriveCtrl and leave it running. Do not exit it.
You should see a lower floor when you try aticonfig --odgc and aticonfig will let you set a lower clock. That's my experience.
|
|
|
I see. So you'd suggest poclmb instead? or?
Another thing - and pardon me for not being up to date - , the two pools (EU and US) are not joined in any way, right?
Right. And Luke answered your question about "waiting for work" thing a few posts up.
|
|
|
The first thing to realize is that this crew immediately assumes the Government/Anonymous/Martians are attacking/controlling/eating the pools/Tycho/bitcoins.
.... a lot of the time... when a pool goes down....
It's because your run-of-the-mill service outage. Nameservers being out of date, connections being lost, "oops" in the code, service provider accidentally blocked port 8337, etc, etc.
I especially enjoy when someone claims DDoS on a pool myself and others are currently active and using with no issues. That's brilliant.
|
|
|
#!/bin/bash while [ 1 ]; do echo "<html><head><meta http-equiv=\"refresh\" content=\"5\"/><body>$(aticonfig --adapter=all --odgt | sed ':a;N;$!ba;s/\n/<br\/>/g')<body></html>" | nc -l 9001 done
I realized we have a lot of new Linux users out there, so here's a simple script to keep you away from the GUI. Put the above in a file, chmod u+x it, run it. That's it. Connect with a browser to whatever the host is, port 9001. Requires netcat (nc). Feel free to replace "aticonfig --adapter=all --odgt" with whatever you like - use tee to redirect your miner output or whatever you want to a file, then push that through netcat. Let's say you've got a script called startminer.sh that starts poclbm (or whatever). You can use the following to start it: startminer.sh | tee -a minerlog.log Then use - #!/bin/bash while [ 1 ]; do echo "<html><head><meta http-equiv=\"refresh\" content=\"5\"/><body>$(sed ':a;N;$!ba;s/\n/<br\/>/g' < minerlog.log)<body></html>" | nc -l 9001 done
For ssh sessions, don't forget that happy watch command - watch -n 1 aticonfig --adapter=all --odgt Will continuously print to the screen the output from aticonfig every 1 second. And finally, if you're using eligius, don't forget there are both EU and US servers. I haven't been able to test the following script... eligius servers haven't gone down. But the idea is if you had a line in your /etc/hosts like: 173.242.112.67 eligius.mining And pointed a miner at http://eligius.mining:8337 while the script below ran..... if the hashrate for that server hit 0 or was otherwise unreachable it should "failover" to the other server. #!/bin/bash
EUHTTP="http://eligius.st/~luke-jr/raw/eu/hashrate.txt" USHTTP="http://eligius.st/~luke-jr/raw/us/hashrate.txt"
EUIP="85.25.78.8" USIP="173.242.112.67"
export EUIP export USIP
CURRENT=$USHTTP
echo $EUIP echo $USIP
while [ 1 == 1 ]; do echo "Checking..."
HR=$(wget -q -O - $CURRENT)
if [ -z "$HR" ] then HR="0" fi
echo "Hashrate $HR"
if [ $CURRENT == $USHTTP ] then if [ $HR == 0 ] then cp /etc/hosts /tmp/hosts sed s/$USIP/$EUIP/ < /tmp/hosts > /etc/hosts echo "Switched to Europe." CURRENT=$EUHTTP fi elif [ $CURRENT == $EUHTTP ] then if [ $HR == 0 ] then cp /etc/hosts /tmp/hosts sed s/$EUIP/$USIP/ < /tmp/hosts > /etc/hosts echo "Switched to US." CURRENT=$USHTTP fi fi sleep 5m done
|
|
|
I get the idle notification with phoenix every once in a while, but less than a second later I've already submitted work and had it accepted....
Not sure why. Maybe it's something flakey with phoenix? It doesn't seem to affect my overall hashrate, though.
|
|
|
have you used HTTPS ? someone may be sniffing your traffic...
That's the best part of St. Eligius - no stupid registration, no password data, no compromise, no sending my email and personal data to random people on the internets...... I can paste my St. Eligius connection string on a billboard and the worst you can do is contribute to me.
|
|
|
For threads like this - MORE DDoS on Deepbit? http://forum.bitcoin.org/index.php?topic=15318.0I bet there's a guy . . . http://forum.bitcoin.org/index.php?topic=16030.0Anyone have an explanation for the growing "anonymous" pool of miners? http://forum.bitcoin.org/index.php?topic=15761.0* No offense to the thread starts, of course, but the content of the thread quickly shifted into entertainment. .... not to mention most of these belong elsewhere, anyway - such as Bitcoin Discussion. Alternate names for the Entertainment subforum could include: "I think this is funny and someone will take it seriously" "CONSPIRACY THEORY - Starring Mel Gibson and Patrick Stewart" "Wildly illogical rants about Deepbit aka 51% Conspiracy aka The Government is Double-spending by Using Mind Rays on Tycho" "Complexity Is Directly Related to Price per Coin" "Proselytize KWh" aka "Cult of the Kill-a-Watt" aka "Paying $0 for CPU mining - you can't explain that" aka "You'll get fired for distributed processing projects on PCs that aren't in your home!" "I think my pool is cheating me even though the statistics and costs are clearly printed somewhere" "OMG a young market is shifting!" "I just got here - am I rich yet?" .... and finally - "I have absolutely not professional I.T. experience and have been alive for fewer years than Linux but it's pretty much the facts that I know better than you - PS who has infrastructure larger than a few PCs in their bedroom?"
|
|
|
I just moved to this pool from btcmine, hope it goes well I'm sure it will! St. Eligius is the easiest pool to get running on. There's a hiccup now and then - but no worse than Slush and Deepbit going batty. On average, I've got about 6 coins in 6 days with an average of somewhere near 350MHs (I think, not including downtime). I usually do burn-in and other tests on St. Eligius, so that 6 days includes roughly 2 whole days of downtime when I was tweaking/refitting cards in one of my miners.
|
|
|
|