Also, my bad, the future firmware update is not about SPI frequency, it's about configuring the PLL settings, which is directly correlated with the hashing speed as you are directly altering the clock speed of the die.
excellent. any ETA for the new fw? |
|
|
|
Thanks.
How many of these should I see in the list?
323 root 2148 S udhcpc -b -x hostname Jupiter-1 eth0
14003 root 2152 S grep dhcp
29312 root 2148 S udhcpc -b -x hostname Jupiter-1 eth0
I think that one is enough. you should kill the one with the lower pid. But i'd prefer if you check it with someone else that is using the dhcp network settings. usually there's only one dhclient for each network interface. If I were you I would simply reboot the miner and check how many udhcpc will be running. but i'd dare to say that you could do this the next time a network glitch appear, this way you could use the downtime to test this solution without wasting mining time. Thank you. I posted my problem at my ISP forum and someone suggested a possible solution where Static IP addresses would not get changed by the router - I should set Static IP addresses that are not in the DHCP pool - anything below .64: https://i.imgur.com/3giMtIs.pngso you have set a static IP that belongs to your dhcp address pool, right? the fact is that if you did not have any dhcp client hanging around there's no way that the BBB let the dhcp server change its own static ip address. now that I've checked better /www/pages/cgi-bin/network.cgi I've found that there's also an avhai deamon involved... |
|
|
|
Thanks.
How many of these should I see in the list?
323 root 2148 S udhcpc -b -x hostname Jupiter-1 eth0
14003 root 2152 S grep dhcp
29312 root 2148 S udhcpc -b -x hostname Jupiter-1 eth0
I think that one is enough. you should kill the one with the lower pid. But i'd prefer if you check it with someone else that is using the dhcp network settings. usually there's only one dhclient for each network interface. If I were you I would simply reboot the miner and check how many udhcpc will be running. but i'd dare to say that you could do this the next time a network glitch appear, this way you could use the downtime to test this solution without wasting mining time. |
|
|
|
I've to say that the miners' network stack is not the best their part, more than once I've found two or more dhclients running on the miner side. For me, switching to a manual network configuration have solved all the problems and glitches that seem related to the network stack, so far.
What do you mean when you say "manual network configuration"? Static IP addresses setup at the miner? or Reserved IP addresses setup at the router like me? The former (static ip at the miner) I tried that at first, but my router kept changing the IP addresses randomly. I think this is due so a stale "udhcpc" (dhcpclient) floating around. If my theory is correct the problem will be fixed by a reboot or by killing the stale dhcpclient process (login via ssh; ps | grep dhcp; kill <stale_process_pid>). |
|
|
|
I've to say that the miners' network stack is not the best their part, more than once I've found two or more dhclients running on the miner side. For me, switching to a manual network configuration have solved all the problems and glitches that seem related to the network stack, so far.
What do you mean when you say "manual network configuration"? Static IP addresses setup at the miner? or Reserved IP addresses setup at the router like me? The former (static ip at the miner) |
|
|
|
Just clicked Apply in each miner Networking Tab and the FREE changed to USED in the DHCP table: https://i.imgur.com/em3yQah.pngThat was probably the cause, but I have no idea why. I'm glad you've found a solution. I've to say that the miners' network stack is not the best their part, more than once I've found two or more dhclients running on the miner side. For me, switching to a manual network configuration have solved all the problems and glitches that seem related to the network stack, so far. I was not implying that your current switch isn't good, far from that. It was just a piece of advice that save me a lot of troubleshooting in the past, the cheapest route is not the safer one, usually. |
|
|
|
Guys, I need you help with the weirdest problem: Since yesterday my miners have started to disappear from my local network all of a sudden and that happens too often now. All my miners are connected to a TP-Link TL-SF1008D 8-Port 10/100Mbps Unmanaged Desktop Switch, which is then connected to my router Technicolor TG582n (given by ISP): https://i.imgur.com/8HmLmKO.pngWhen the problems happens the blinking light at the back of the router, where the switch is connected has gone off, yet the all 7 lights (6 miners + incoming cable) are ON. Until yesterday I wasn't sure what is the problem, but now I know: when I unplug the switch from the router and plug it back in = they appear back in the network. Additionally this morning my own PC reported network conflict: some other machine on the network got its IP too, which never happened. + I have setup the DHCP table manually. The 101 to 106 are the miners. Before it said state USED and now it is free, but they are hashing. https://i.imgur.com/GPiLxxb.pngDo you have any experience with bad switches? Any ideas what is causing the issue? Thank you. if you could, just go down the road and buy a new switch, hopefully it is the cause of the problem. A 8 port switch cost almost nothing. buy a good one while at it. |
|
|
|
I love high-inflation countries that don't have widespread Btc adoption yet.
Japan, Argentina, and now India. Potentially huge positive development. They already love gold, wait til average Indian hears about Btc - or if they start gifting Btc instead of gold for their wedding season. Choo choo.
sorry but "high-infaltion" and "Japan" can't be on the same phrase  http://www.inflation.eu/inflation-rates/japan/historic-inflation/cpi-inflation-japan.aspx |
|
|
|
wow. every time KnC hits the market this kind of growth happens. Anyone have an an idea of in what measure other asic vendors (bitmain, bitfury, avalon 2nd gen) contributed to this last jump? I don't think there were a lot of KNC's shipped this difficulty. Most of the November batch would have been shipped during the previous difficulty. I suspect that most of the TH added this difficulty were probably from private mining operations. you were right. this is an example $2 million+ bitcoin mining facility launched in Iceland . /sarcasm I'm not blaiming KnC for having sold so much hashrate to a single entity. I'm just wondering how many others had the same idea and were able to stash enough hash power to make their project worthty. |
|
|
|
bitstamp is an amazing exchange. Only thing it doesn't have is an auto log out feature. This would help with security....
Did they implement stop loss order? |
|
|
|
wow. every time KnC hits the market this kind of growth happens. Anyone have an an idea of in what measure other asic vendors (bitmain, bitfury, avalon 2nd gen) contributed to this last jump? |
|
|
|
|
I'm a newbie trading wise, so this my dumb question: who's buying all those cheap coin? Because if someone is selling someone else has to buy, no?
|
|
|
|
Anyone knows if the Oct controller board is compatible with Nov type ASIC boards?
It is. The only difference is the firmware. Flash 0.99-E onto it and you should be all set. Weren't we warned not to try that? I'm sure they have re-designed the pcb asic board for nov miners, dunno if they did the same for the main controller board. |
|
|
|
Anyone knows if the Oct controller board is compatible with Nov type ASIC boards?
It is. The only difference is the firmware. Flash 0.99-E onto it and you should be all set. Pay attention as of now you can't mix october with nov boards on the same controller board edit: fix a thinko |
|
|
|
friendcat is bad for mining
Friedcat's miners could be not the best deal for a wanna be miner, but he's definitely good at mining. |
|
|
|
You do realize they started 20nm about three months ago?... and they said Q1/Q2.
Plenty of cushion IMHO
Doesn't matter if they started 10 years ago. A huge limitation to actually meeting such a date is TSMC's ability to deliver. They haven't yet ramped up 20nm production! TSMC has only taped-out a handful of 20nm designs so far. TSMC production is going to be very limited in Q1. If TSMC does get in a crunch, who are they going to deliver wafers to first, Apple or KnC? Maybe they will use knc as a further way to test their production line before going at full speed  |
|
|
|
What have you sold 12 of, Jupiters? or Neptunes?
One morning I got a UPS confirmation, then they arrived on the Monday.
I've sold 12 Neptunes... I was surprised to see it myself, I guess it paid off to have the ad in my signature, plus I paid for advertising on an online classifieds called 'Gumtree' in Australia. That was best $27.95 I ever spent. I didn't offer free hash, but I offered free support, setup and maintenance for 12 months. Plus I bought one myself, but I didn't use my reseller link for myself... I forgot with all the excitement when they went on sale I couldn't be happier if I get a bonus  that'll make 2 Neptunes coming my way kudos merv. if it was up to me I would give you two neptune as bonus just for how much I like your avatar |
|
|
|
A show of hands, please? Does anyone here know someone who's overclocking their machine? I know none of you folks, myself included, would ever void the warranty on such a valuable machine, but OC'ing appears to offer great promise for October machines and machines with die-0 issues. One friend on mine has a October Jupiter with two die 0 boards that was pushing 495Ghs, but it skyrocketed to over 560Ghs when she edited the cgminer.sh file to look like this:
if [ $p -eq 2 ]
then
cmd=$(printf "0x86,0x%02X,0x02,0x11" $c)
if [ $c -eq 1 ]
then
cmd=$(printf "0x86,0x%02X,0x03,0xE2" $c)
fi Is my friend's experience unique? Both bad die-0 boards are still silent, yet her hashrate is as high as my 'good' jupiter with no bad dies. Her wattage-per-die is right at 40W, so this is pushing the limits of the specifications of the VRMs, but we added extra cooling fans inside her machine to keep the temps down low to compensate. No hiccups yet! take your time and read all this thread, especially tolip_wen posts. you'll find the answer to your question. |
|
|
|
No please stop pushing the network up for christ sake. Didn't KnC already do enough damage?
good one. you made me laugh. you've bought 3 jups + 6 expansion boards already and maybe also a few neptunes. I mean, you too are contributing to the damage you're ascribing to knc. no? And don't you think that HF/CT/BA would do as much as damage as they can to the network rate, despite to what KnC will do? |
|
|
|
Neptunes are down to 32/1200 left up for sell.
Seems they will have sold out 1200 x 12,999.00 = WOW
so they have more that what is needed to cover NRE cost. |
|
|
|
|
Show Posts
