cet
Member
 Offline
Activity: 100
Merit: 10
|
 |
September 30, 2013, 08:21:39 AM |
|
Does anyone have some good pointers for setting up a network file server to point several rPIs at? I've had a couple of memory cards fail already and I'd like to move to something more dependable.
/cet
|
|
|
|
|
Jutarul
Donator
Legendary
Offline
Activity: 994
Merit: 1000
|
|
September 30, 2013, 08:31:43 AM |
|
Does anyone have some good pointers for setting up a network file server to point several rPIs at? I've had a couple of memory cards fail already and I'd like to move to something more dependable.
/cet
just export a separate directory for each raspberry pi and associate it with the IP address each individual pi is booted with (as specified in the cmdline.txt). Then your /etc/exports file should look similar to this: # /etc/exports: the access control list for filesystems which may be exported
# to NFS clients. See exports(5).
#
# Example for NFSv2 and NFSv3:
# /srv/homes hostname1(rw,sync,no_subtree_check) hostname2(ro,sync,no_subtree_check)
#
# Example for NFSv4:
# /srv/nfs4 gss/krb5i(rw,sync,fsid=0,crossmnt,no_subtree_check)
# /srv/nfs4/homes gss/krb5i(rw,sync,no_subtree_check)
#
/var/lib/tftpboot/raspberry/192.168.1.249 192.168.1.249(rw,no_root_squash,async,insecure,no_subtree_check)
/var/lib/tftpboot/raspberry/192.168.1.248 192.168.1.248(rw,no_root_squash,async,insecure,no_subtree_check)
Before, you should have copied the rootfs to the server, e.g. by mounting directly the SD card: mount /dev/sdb2 /mnt/sdb2 mkdir -p /var/lib/tftpboot/raspberry/template cp -ax /mnt/sdb2/* /var/lib/tftpboot/raspberry/template and populate each new client: mkdir /var/lib/tftpboot/raspberry/192.168.1.249 cp -ax /var/lib/tftpboot/raspberry/template/* /var/lib/tftpboot/raspberry/192.168.1.249 The IPs dependent on your particular network configuration. As long as the number of clients is low, the network and the server should be able to handle it. If somebody has access to >40 pi's, it would be worth some benchmarking to find out how large this can scale. |
|
|
|
|
jlsminingcorp
|
|
September 30, 2013, 08:33:45 AM |
|
I'm curious to hear more details from people who have fried pi's. Are you using older or inexpensive PSUs that are plugged directly into the wall or a high quality PSU plugged into an UPS?
Brand new (relatively) high quality 80%-efficiency with UPS. I'm confident that my PSU is generating stable 12V for the mboard. I measured the 5v power coming out of the m-board pins and going into the RPi. It's weird... first it's at 5.39V but it keeps increasing every couple seconds. In any case, 5.39V is beyond the maximum (5.25V) What I'm finding is that the polyfuse F3 (near the micro-usb connector) gets really really hot - finger burning. Is yours like that? Doesn't a hot polyfuse mean that it's blown or in the process of blowing? What's the resistance of the fuse now (see http://www.raspberrypi.org/phpBB3/viewtopic.php?t=19033 and http://elinux.org/RPi_Hardware for some related discussion)? These overcooked RPis are a bit of a worry though aren't they? It sounds like it's happening with a range of different PSUs (including some high-end units), which suggests that it's not necessarily a PSU problem. Is it possible that the way that some m-boards are stepping down the voltage from 12V to 5V to supply the RPi is causing the problem? Apologies for being an armchair EE here (I don't have a lot of experience in the area), perhaps one of the more experienced EEs on the forum can give more insight (and correct my errors ;-)). |
|
|
|
|
xstr8guy
|
|
September 30, 2013, 08:41:14 AM |
|
Lol, being a PSU snob has come 'round and bit me in the ass! Two pi's and 2 SD cards now appear to be faulty and or corrupted. The pi's boot up but then they go into a never ending cycle of debugging gibberish. Where do I start?! I know, go back and read... doing that now. But if anyone can point me to a shortcut, that would be cool.  Don't worry, this may not be related to power at all, the card/os may just have got corrupted, it happens. You were being a bit of a snob though ;-). Problem resolved and lesson learned. I just had to recreate the img file on the SD and reconfigure. Took 30 minutes and most of that time was spent waiting for img file to download and copy to the SD card. Panic attack aborted! And the lessoned learned is... don't be a dick! Or the pi gods will get you. Lol. |
|
|
|
|
|
jlsminingcorp
|
|
September 30, 2013, 08:45:54 AM |
|
Lol, being a PSU snob has come 'round and bit me in the ass! Two pi's and 2 SD cards now appear to be faulty and or corrupted. The pi's boot up but then they go into a never ending cycle of debugging gibberish. Where do I start?! I know, go back and read... doing that now. But if anyone can point me to a shortcut, that would be cool. Don't worry, this may not be related to power at all, the card/os may just have got corrupted, it happens. You were being a bit of a snob though ;-). ok Im freaking out a little bit....are there problems with the raspberry Pis on the current M2 boards that are expected for october delivery? Does using a ribbon cable seem to address these issues (someone said that the interface used to connect the PI to the M-Board is not the best way to power the pi is there is no overvoltage regulation? Its sounding like the input from the mboard eventually gets high like around 5.4V when it should be staying around 5v even? I'll email their support to be sure...but I was wondering if you guys could put my fears to ease beforehand? Does having a different amount of H-boards connected make a difference to longevity? I'm also planning to install into a case made by spotswood will the fans used hopefully mitigate the problem with the pi? Should I consider buying a 2nd pi and sd card to make a drop in backup? Have you guys done the same if the answer to that was yes? I've already dropped a substantial amt of coin on a miner from them but in retrospect I wish i just went along with cex.io now that i see its on demand. +1, would be good to get some input from BFSB/megabigpower on this. I don't think that cooling will help if the m-board supplies too high a voltage to the pi. Also, just using a ribbon cable to connect to the GPIO pins on the pi wont help. You would have to cut the power wires on the ribbon cable and then supply power through the usb power socket on the pi to bypass the 5V supply from the m-board. Presumably you shouldn't do this unless you know what you're doing. We've got a backup pi + SD cards because of all of this. For a few $ it makes sense having them around I guess. Even if the power issue is fixed the SD cards/os die pretty easily and so it's a good idea to have a spare or two with pre-flashed images on them. |
|
|
|
|
jlsminingcorp
|
|
September 30, 2013, 08:50:33 AM |
|
Lol, being a PSU snob has come 'round and bit me in the ass! Two pi's and 2 SD cards now appear to be faulty and or corrupted. The pi's boot up but then they go into a never ending cycle of debugging gibberish. Where do I start?! I know, go back and read... doing that now. But if anyone can point me to a shortcut, that would be cool. Don't worry, this may not be related to power at all, the card/os may just have got corrupted, it happens. You were being a bit of a snob though ;-). Problem resolved and lesson learned. I just had to recreate the img file on the SD and reconfigure. Took 30 minutes and most of that time was spent waiting for img file to download and copy to the SD card. Panic attack aborted! And the lessoned learned is... don't be a dick! Or the pi gods will get you. Lol. Good to hear that you're back up and running . It's also encouraging that your RPis weren't toasted, so this may not be related to the power issue. Are you getting stable 5V output from your m-boards by the way or do you see the worrying voltages that dracore has seen? |
|
|
|
|
xstr8guy
|
|
September 30, 2013, 09:53:04 AM |
|
Lol, being a PSU snob has come 'round and bit me in the ass! Two pi's and 2 SD cards now appear to be faulty and or corrupted. The pi's boot up but then they go into a never ending cycle of debugging gibberish. Where do I start?! I know, go back and read... doing that now. But if anyone can point me to a shortcut, that would be cool. Don't worry, this may not be related to power at all, the card/os may just have got corrupted, it happens. You were being a bit of a snob though ;-). Problem resolved and lesson learned. I just had to recreate the img file on the SD and reconfigure. Took 30 minutes and most of that time was spent waiting for img file to download and copy to the SD card. Panic attack aborted! And the lessoned learned is... don't be a dick! Or the pi gods will get you. Lol. Good to hear that you're back up and running . It's also encouraging that your RPis weren't toasted, so this may not be related to the power issue. Are you getting stable 5V output from your m-boards by the way or do you see the worrying voltages that dracore has seen? I don't have a clue how to check voltages. Flying blind here! Lol. And yes everyone, have an extra pi and a handful of SD cards on hand. And an extra M board probably would be wise. I guess this is what we get for demanding our ASICs be delivered quickly without proper testing.  |
|
|
|
|
af_newbie
Legendary
Offline
Activity: 2702
Merit: 1468
|
|
September 30, 2013, 12:23:19 PM |
|
Looks like KNC came through. Quick, run for cover, cancel all bitfury orders....
Flood of KNCs are coming in October, network hash rate is going to 4PH, diff to 0.5B
|
|
|
|
|
zurg
|
|
September 30, 2013, 12:45:16 PM |
|
Looks like KNC came through. Quick, run for cover, cancel all bitfury orders....
Flood of KNCs are coming in October, network hash rate is going to 4PH, diff to 0.5B
Heck even without KNC, but looking at calculator (the % increase, 30 days 126%,60 days 376%, 90 days 597%) and taking into account pencil-modding new board (EU thread) to 35Gh/s Is not even going to come even close to ROI at $500 per H-board.  My $3500 investment (of 7 H-boards) would leave me $1290 in the hole and that's with 0 electricity costs with me starting to mine 1st of Oct. http://mining.thegenesisblock.com/ |
|
|
|
|
af_newbie
Legendary
Offline
Activity: 2702
Merit: 1468
|
|
September 30, 2013, 02:03:54 PM |
|
Looks like KNC came through. Quick, run for cover, cancel all bitfury orders....
Flood of KNCs are coming in October, network hash rate is going to 4PH, diff to 0.5B
Heck even without KNC, but looking at calculator (the % increase, 30 days 126%,60 days 376%, 90 days 597%) and taking into account pencil-modding new board (EU thread) to 35Gh/s Is not even going to come even close to ROI at $500 per H-board. My $3500 investment (of 7 H-boards) would leave me $1290 in the hole and that's with 0 electricity costs with me starting to mine 1st of Oct. http://mining.thegenesisblock.com/I'm seriously thinking about cancelling my order of 30 boards. At $500 a pop, they don't look too hot even at 35GH/s per board. I had 2 more orders lined up for full kits, but it is hard to justify the 8K sticker at this difficulty progression. And from what I hear from Dave, he is planning to send them out 3rd or 4th week of October, at best. |
|
|
|
|
zurg
|
|
September 30, 2013, 02:12:29 PM |
|
Looks like KNC came through. Quick, run for cover, cancel all bitfury orders....
Flood of KNCs are coming in October, network hash rate is going to 4PH, diff to 0.5B
Heck even without KNC, but looking at calculator (the % increase, 30 days 126%,60 days 376%, 90 days 597%) and taking into account pencil-modding new board (EU thread) to 35Gh/s Is not even going to come even close to ROI at $500 per H-board. My $3500 investment (of 7 H-boards) would leave me $1290 in the hole and that's with 0 electricity costs with me starting to mine 1st of Oct. http://mining.thegenesisblock.com/I'm seriously thinking about cancelling my order of 30 boards. At $500 a pop, they don't look too hot even at 35GH/s per board. I had 2 more orders lined up for full kits, but it is hard to justify the 8K sticker at this difficulty progression. And from what I hear from Dave, he is planning to send them out 3rd or 4th week of October, at best. First October orders are already being received in EU thread. Looks like we're getting screwed a bit on timing by a 3-4 weeks. Dunno.. maybe Dave will clarify. My my calculations right now the boards should be MAX $250 and that would make about 25% ROI (no electric costs), but even that for me is to of a risky proposition at this point. I am definitely not touching any $500 h-boards at this point. |
|
|
|
|
af_newbie
Legendary
Offline
Activity: 2702
Merit: 1468
|
|
September 30, 2013, 02:21:48 PM |
|
First October orders are already being received in EU thread. Looks like we're getting screwed a bit on timing by a 3-4 weeks. Dunno.. maybe Dave will clarify. My my calculations right now the boards should be MAX $250 and that would make about 25% ROI (no electric costs), but even that for me is to of a risky proposition at this point. I am definitely not touching any $500 h-boards at this point. He emailed me on the weekend. 3rd week of October at the earliest. But he said that they will try to send out all October orders in October. I guess his 200TH mine is a priority. If KNC starts really shipping orders in October, that bitfury full Kit becomes 4-5K kit overnight. Then depreciation will depend on the rate KNC adds hashing power to the network. Once people behind 100TH mine project realize that they need to make that 1000TH instead of 200TH mine, then we'll not see October orders until sometime in November/December. |
|
|
|
Dexter770221
Legendary
Offline
Activity: 1029
Merit: 1000
|
|
September 30, 2013, 02:25:07 PM |
|
First October orders are already being received in EU thread. Looks like we're getting screwed a bit on timing by a 3-4 weeks. Dunno.. maybe Dave will clarify. My my calculations right now the boards should be MAX $250 and that would make about 25% ROI (no electric costs), but even that for me is to of a risky proposition at this point. I am definitely not touching any $500 h-boards at this point. He emailed me on the weekend. 3rd week of October at the earliest. But he said that they will try to send out all October orders in October. I guess his 200TH mine is a priority. If KNC starts really shipping orders in October, that bitfury full Kit becomes 4-5K kit overnight. Then depreciation will depend on the rate KNC adds hashing power to the network. Once people behind 100TH mine project realize that they need to make that 1000TH instead of 200TH mine, then we'll not see October orders until sometime in November/December. Remember that Dave is just a resseler. Tytus is a producer of the boards and founder of 100TH mine. MBP is a host of 200TH mine and is obligated to deliver this hash power as fast as possible. That means no stock for resseling. Punin only resseling thats why he has stock and already sends October orders. |
Under development Modular UPGRADEABLE Miner (MUM). Looking for investors.
Changing one PCB with screwdriver and you have brand new miner in hand... Plug&Play, scalable from one module to thousands.
|
|
|
Cablez
Legendary
Offline
Activity: 1400
Merit: 1000
I owe my soul to the Bitcoin code...
|
|
September 30, 2013, 02:26:24 PM |
|
Yes, these are the constraints that punin doesn't seem to have as he is solely a reseller. Hence faster ship times.
|
Tired of substandard power distribution in your ASIC setup??? Chris' Custom Cablez will get you sorted out right! No job too hard so PM me for a quote
Check my products or ask a question here: https://bitcointalk.org/index.php?topic=74397.0
|
|
|
af_newbie
Legendary
Offline
Activity: 2702
Merit: 1468
|
|
September 30, 2013, 02:30:44 PM |
|
Yes, these are the constraints that punin doesn't seem to have as he is solely a reseller. Hence faster ship times.
Yes, looks like we need another US/NA reseller that is not bound by other commitments. |
|
|
|
|
zurg
|
|
September 30, 2013, 02:34:22 PM |
|
This kinda puts us in US under big disadvantage.
We pay roughly (i haven't checked in a while) same as EU customers yet we get same product 3-4 weeks after EU. And we all know what that means.
|
|
|
|
|
dben428
Member
Offline
Activity: 106
Merit: 10
|
|
September 30, 2013, 03:24:14 PM |
|
I had posted this in the support thread. I'm hoping to get a quicker response here.
I had been using BTC Guild since I had gotten my miner, and everything was fine up until last night when BTC Guild got hit with a DDoS attack. Since then, I am not able to connect to BTC Guild or any other pool (I've tried 50BTC, Eclipse, and Bitminter). Does anyone have any suggestions? Thanks.
|
|
|
|
|
af_newbie
Legendary
Offline
Activity: 2702
Merit: 1468
|
|
September 30, 2013, 03:33:04 PM |
|
I had posted this in the support thread. I'm hoping to get a quicker response here.
I had been using BTC Guild since I had gotten my miner, and everything was fine up until last night when BTC Guild got hit with a DDoS attack. Since then, I am not able to connect to BTC Guild or any other pool (I've tried 50BTC, Eclipse, and Bitminter). Does anyone have any suggestions? Thanks.
Eclipse is working fine for me. See if you can ping, telnet to the pool from your miner. Ethernet LEDs are ok? If not, re-image the card and try again. |
|
|
|
|
tom99
|
|
September 30, 2013, 03:39:33 PM |
|
I had posted this in the support thread. I'm hoping to get a quicker response here.
I had been using BTC Guild since I had gotten my miner, and everything was fine up until last night when BTC Guild got hit with a DDoS attack. Since then, I am not able to connect to BTC Guild or any other pool (I've tried 50BTC, Eclipse, and Bitminter). Does anyone have any suggestions? Thanks.
EMC is working for me too and are you using port: 3333 for first one |
|
|
|
|
cet
Member
Offline
Activity: 100
Merit: 10
|
|
September 30, 2013, 03:47:22 PM |
|
I had posted this in the support thread. I'm hoping to get a quicker response here.
I had been using BTC Guild since I had gotten my miner, and everything was fine up until last night when BTC Guild got hit with a DDoS attack. Since then, I am not able to connect to BTC Guild or any other pool (I've tried 50BTC, Eclipse, and Bitminter). Does anyone have any suggestions? Thanks.
bitminter has been running just fine. /cet |
|
|
|
|
|
Poll
