Topic: ANTMINER C1 Discussion and Support Thread - page 19. (Read 129054 times)

Quote from: soy on March 24, 2015, 06:15:50 PM

A difficulty with my C1#2. I changed the parameters including password and can log in on the gui. When I ssh in, I register the key in putty then try to log in but my password isn't accepted. Neither the new password nor password root are accepted in ssh.

Bothersome because I always like to command halt via ssh before killing power. I found wakeup was less troublesome in earlier miners.

Signed in with admin as password, then changed root's passwd.

Well lost password access via ssh on 2 C1's just now. I knew that when new to sign in via ssh one would need to sign in as root and use admin as the password. So, when my regular password wouldn't work I tried that but that's no go either.

What happened was I put a script in /sbin called cg.sh to restart /etc/init.d/cgminer.sh but of course it would be lost with the next boot so I ran the 'cp -rf /etc/shadow /config/' then on the next boot I tried to login to run the script but can't get in.

I have tried changing the password in the GUI. I can still get into the GUI no problem just can't ssh in.

The good thing with C1's is you could chances are flash the micro sd card and regain ssh access.

Can't get the cover off without breaking down its installation. There's only a subnet accessible in the back room. Resetting puts it back on the .1.0 network. I can leave it in place with circ water cooling and reset while hooked to a router not connected to the internet. Big discussion in S5 forum about how loss of internet gets overheating. At least the C1's cooling continues after internet break. But it's hashing perfectly and I have the GUI.

I'd like to know if this failure is a built-in safeguard of some kind. Putting a short script in /sbin/, which is on the path unlike /usr/sbin/ then saving with copy -rf /etc/shadow to /config/ then rebooting and losing access via ssh. Tried root/mypassword, root/admin, admin/admin and none get me in. If there were a cron entry like the S3's I might try putting in passwd root and running it once but no cron on C1's. ..and now that I think of it - it wouldn't work as it comes back with the enter new password request

What ill it be breaking? I know on mine if I unplug power I can open it up without needing to disconnect the fluid. But I am also using a molex to 3 pin adapter so no fan are powered inside the miner.

I have my miners on inexpensive, 4 shelf, wheeled carts from Walmart's. Supplies up top and miner below. S3's could slide in but C1's I had to cut out a bar to increase access. Shelves are too close to remove a cover in place. For the S3's the mobility is good as I can just spin the cart if I reverse a window fan but of course the C1's don't have much latitude with the radiators mounted in another window.

notlist3d

legendary

Activity: 1456

Merit: 1000

Quote from: notlist3d on March 24, 2015, 05:50:04 PM

Quote from: notlist3d on March 24, 2015, 05:50:04 PM

A difficulty with my C1#2. I changed the parameters including password and can log in on the gui. When I ssh in, I register the key in putty then try to log in but my password isn't accepted. Neither the new password nor password root are accepted in ssh.

Bothersome because I always like to command halt via ssh before killing power. I found wakeup was less troublesome in earlier miners.

Signed in with admin as password, then changed root's passwd.

Well lost password access via ssh on 2 C1's just now. I knew that when new to sign in via ssh one would need to sign in as root and use admin as the password. So, when my regular password wouldn't work I tried that but that's no go either.

What happened was I put a script in /sbin called cg.sh to restart /etc/init.d/cgminer.sh but of course it would be lost with the next boot so I ran the 'cp -rf /etc/shadow /config/' then on the next boot I tried to login to run the script but can't get in.

I have tried changing the password in the GUI. I can still get into the GUI no problem just can't ssh in.

The good thing with C1's is you could chances are flash the micro sd card and regain ssh access.

Can't get the cover off without breaking down its installation. There's only a subnet accessible in the back room. Resetting puts it back on the .1.0 network. I can leave it in place with circ water cooling and reset while hooked to a router not connected to the internet. Big discussion in S5 forum about how loss of internet gets overheating. At least the C1's cooling continues after internet break. But it's hashing perfectly and I have the GUI.

I'd like to know if this failure is a built-in safeguard of some kind. Putting a short script in /sbin/, which is on the path unlike /usr/sbin/ then saving with copy -rf /etc/shadow to /config/ then rebooting and losing access via ssh. Tried root/mypassword, root/admin, admin/admin and none get me in. If there were a cron entry like the S3's I might try putting in passwd root and running it once but no cron on C1's. ..and now that I think of it - it wouldn't work as it comes back with the enter new password request

What ill it be breaking? I know on mine if I unplug power I can open it up without needing to disconnect the fluid. But I am also using a molex to 3 pin adapter so no fan are powered inside the miner.

soy

legendary

Activity: 1428

Merit: 1013

A difficulty with my C1#2. I changed the parameters including password and can log in on the gui. When I ssh in, I register the key in putty then try to log in but my password isn't accepted. Neither the new password nor password root are accepted in ssh.

Bothersome because I always like to command halt via ssh before killing power. I found wakeup was less troublesome in earlier miners.

Signed in with admin as password, then changed root's passwd.

Well lost password access via ssh on 2 C1's just now. I knew that when new to sign in via ssh one would need to sign in as root and use admin as the password. So, when my regular password wouldn't work I tried that but that's no go either.

What happened was I put a script in /sbin called cg.sh to restart /etc/init.d/cgminer.sh but of course it would be lost with the next boot so I ran the 'cp -rf /etc/shadow /config/' then on the next boot I tried to login to run the script but can't get in.

I have tried changing the password in the GUI. I can still get into the GUI no problem just can't ssh in.

The good thing with C1's is you could chances are flash the micro sd card and regain ssh access.

Can't get the cover off without breaking down its installation. There's only a subnet accessible in the back room. Resetting puts it back on the .1.0 network. I can leave it in place with circ water cooling and reset while hooked to a router not connected to the internet. Big discussion in S5 forum about how loss of internet gets overheating. At least the C1's cooling continues after internet break. But it's hashing perfectly and I have the GUI.

I'd like to know if this failure is a built-in safeguard of some kind. Putting a short script in /sbin/, which is on the path unlike /usr/sbin/ then saving with copy -rf /etc/shadow to /config/ then rebooting and losing access via ssh. Tried root/mypassword, root/admin, admin/admin and none get me in. If there were a cron entry like the S3's I might try putting in passwd root and running it once but no cron on C1's. ..and now that I think of it - it wouldn't work as it comes back with the enter new password request

notlist3d

legendary

Activity: 1456

Merit: 1000

A difficulty with my C1#2. I changed the parameters including password and can log in on the gui. When I ssh in, I register the key in putty then try to log in but my password isn't accepted. Neither the new password nor password root are accepted in ssh.

Bothersome because I always like to command halt via ssh before killing power. I found wakeup was less troublesome in earlier miners.

Signed in with admin as password, then changed root's passwd.

Well lost password access via ssh on 2 C1's just now. I knew that when new to sign in via ssh one would need to sign in as root and use admin as the password. So, when my regular password wouldn't work I tried that but that's no go either.

What happened was I put a script in /sbin called cg.sh to restart /etc/init.d/cgminer.sh but of course it would be lost with the next boot so I ran the 'cp -rf /etc/shadow /config/' then on the next boot I tried to login to run the script but can't get in.

I have tried changing the password in the GUI. I can still get into the GUI no problem just can't ssh in.

The good thing with C1's is you could chances are flash the micro sd card and regain ssh access.

soy

legendary

Activity: 1428

Merit: 1013

Quote from: soy on March 24, 2015, 12:05:04 PM

A difficulty with my C1#2. I changed the parameters including password and can log in on the gui. When I ssh in, I register the key in putty then try to log in but my password isn't accepted. Neither the new password nor password root are accepted in ssh.

Bothersome because I always like to command halt via ssh before killing power. I found wakeup was less troublesome in earlier miners.

Signed in with admin as password, then changed root's passwd.

Well lost password access via ssh on 2 C1's just now. I knew that when new to sign in via ssh one would need to sign in as root and use admin as the password. So, when my regular password wouldn't work I tried that but that's no go either.

What happened was I put a script in /sbin called cg.sh to restart /etc/init.d/cgminer.sh but of course it would be lost with the next boot so I ran the 'cp -rf /etc/shadow /config/' then on the next boot I tried to login to run the script but can't get in.

I have tried changing the password in the GUI. I can still get into the GUI no problem just can't ssh in.

soy

legendary

Activity: 1428

Merit: 1013

Perplexing. The JB Weld quite cured I put the after market pump back in the loop miner loop after running an hour on a closed loop without a leak. Once in the miners' loop I got a leak at the out port. Figuring it might be the hose I replaced the hose at that port. It leaks again. I suspect this new leak at the seam between the ports or one side of a port tube. I noticed the halves weren't mated perfectly and lightly filed down the ridge along the port tube. After removing the pump from the circuit, with a short tube on one port and my thumb on another I submerged it in a sink and applied air pressure expecting to see bubbles - no bubbles. Perhaps the out port needs the downward pressure of a tube filled with water to leak. ...Yep, applying downward pressure on the out port while submerged and applying air pressure shows a leak along its top interface with the box.

soy

legendary

Activity: 1428

Merit: 1013

The stock pump seems to have thicker walls than the after market Syscooling pump, the one that broke.

After watching immersion cooling in Novec I know I'd love to do the same except it's so expensive. Naturally my thoughts go to immersion in R134a but at room temperature there'd be 89.9psi so something like 1" Lexan without thru holes would be needed with a steel top for thru connections of hoses and conductors and who can afford the 1" Lexan while considering ROI. Could you imagine subjecting a thin-walled after market Syscooling pump to hundreds of psi? Or a steel tank with LED lighting and a videocam.

notlist3d

legendary

Activity: 1456

Merit: 1000

It seems clear Bitmain lost confidence in Syscooling. The fault as I see it may be entirely in Syscooling's bonding process although the wall thickness is likely an element. Removing the tank leaving only a small section to break off would mean only that small section was fully bonded to the bottom.

Honestly it is not a bad thing on losing confidence in them. I hope they never partner again personally.

soy

legendary

Activity: 1428

Merit: 1013

It seems clear Bitmain lost confidence in Syscooling. The fault as I see it may be entirely in Syscooling's bonding process although the wall thickness is likely an element. Removing the tank leaving only a small section to break off would mean only that small section was fully bonded to the bottom.

notlist3d

legendary

Activity: 1456

Merit: 1000

I should have 2 of the syscooling pumps in my parts that were shipped as replacement. I am thinking about selling them, but a little afraid of selling and they break on someone.

On a side note I know long ago it sounded like I was going to get my money back since pumps were lost in mail. But since one showed up in a month and other showed up over a month later I did not receive any credit/charge back. Just incase anyone cared. I could have lied and got it, but was not worth losing my honesty over.

soy

legendary

Activity: 1428

Merit: 1013

Quote from: soy on March 23, 2015, 10:21:52 AM

5 out of 5 pumps developed leaks this week. seems the glue on the reservoir housing fails under temperature

My stock pump is okay but I chose to position it between the radiator cool port and the miner reasoning lower temperature means less wear.

The second pump, different model manufactured by Syscooling, only developed the leak when I tapped on it with a screwdriver to bring bubbles to the surface. After laying down a line of glue along the back seam and letting it dry some hours, it wasn't completely dry except on its surface. I pressed on the dried surface and saw a couple of dots of wet glue pass to the inside where the leaks had developed.

Pump repair was a failure. Will try to remove the tank and re-glue.

Put the second radiator after the stock stock and before the pump feeding C1#1. C1#1 is now the cooler of the two C1's. Previously C2 was cooler, cooled with the condenser.

Took the pump, gripped the tank snuggly, pulled, the tank came free of the bottom.

-------------------------------------------------------------------------------

I thought the top simply separated from the bottom but it turns out it wasn't quite that. The edge of the clear plastic seemed jagged and in a back corner a considerable amount of the clear plastic remained glued to the black bottom, broken off about an eighth inch high. Perhaps the top wasn't glued but welded to the bottom with a chemical that liquified the edge for a bit and allowed them to bond. I went with JB Weld instead of 5 minute epoxy, putting a bead along both surfaces and mating. Tomorrow I'll run a tube from the in port to the out port and run to check for leaks. Leaks near the electronics are disconcerting.

thedreamer

legendary

Activity: 1694

Merit: 1002

Go Big or Go Home.....

Why are these pumps so crappy?

I have been building/selling and running watercooled gaming rigs for years with NEVER any leaks on any of them.
Most of them are driven hard and hot, yet no issues.

I wonder if there's a way to plumb in one of the CPU cooling pumps as a closed look system.

soy

legendary

Activity: 1428

Merit: 1013

5 out of 5 pumps developed leaks this week. seems the glue on the reservoir housing fails under temperature

My stock pump is okay but I chose to position it between the radiator cool port and the miner reasoning lower temperature means less wear.

The second pump, different model manufactured by Syscooling, only developed the leak when I tapped on it with a screwdriver to bring bubbles to the surface. After laying down a line of glue along the back seam and letting it dry some hours, it wasn't completely dry except on its surface. I pressed on the dried surface and saw a couple of dots of wet glue pass to the inside where the leaks had developed.

notlist3d

legendary

Activity: 1456

Merit: 1000

5 out of 5 pumps developed leaks this week. seems the glue on the reservoir housing fails under temperature

Ouch hope you get it back up soon. Did they crack or just start leaking out side?

MyRig

donator

Activity: 792

Merit: 510

[email protected]

If your units are still under the warranty, please contact [email protected] with your Bitmaintech.com Order ID and full shipping information with a telephone number for further instruction.