key files are in a different location that a standard linux distro: look in /etc/dropbear for ssh files e.g.
root@antminer-6:~# ls -alt /etc/dropbear/
-rw-r--r-- 1 root root 1214 Apr 10 09:02 authorized_keys
drwxr-xr-x 1 root root 0 Feb 21 14:17 .
-rw------- 1 root root 459 Dec 1 20:28 dropbear_dss_host_key
-rw------- 1 root root 427 Dec 1 20:28 dropbear_rsa_host_key
drwxrwxr-x 1 root root 0 Dec 1 20:28 ..
root@antminer-6:~#
this is a handy bit of info.. :)
thank you
Sorry about that, jjimm_64 -- I'd just assumed you were cutting and pasting your
id_rsa.pub key data into the Antminer web interface. I guess that just became part of my "initial setup ritual" -- unbox, plug in, connect crossover cable, hit the web ui, change the IP/DNS/router/hostname, paste my pubkey in to the box on the Admin page, hit "apply all" and then reconnect it to the LAN as it reboots itself... (Then I do whatever remaining stuff via ssh.) So this explains why you would have problems and I wouldn't -- I was adding my key data using the wimpy, GUI, "hold my hand and do it for me" way...meanwhile, you were going for old fashioned commandline goodness...and i certainly can't fault you for any of that. :)
As jimrome pointed out, it would appear the Luci Admin page dumps its contents into two
authorized_keys files -- or one, and then another is a clone of that one, I'm not sure. (Haven't really studied the Antminer file system setup, and not really looking to do so at the moment. lol) Anyway, my key data (which I pasted into the Luci Admin page) appeared in full, in both places, so you might want to copy your data in both of these also (if it isn't automatically cloned over due to the file overlay magic):
- /etc/dropbear/authorized_keys
- /overlay/etc/dropbear/authorized_keys
Not sure if they're the same inode, since stat isn't installed and I'm too lazy to look for an alternative at the moment -- but, they do both contain the same data:
root@ant0:/# cat /overlay/etc/dropbear/authorized_keys && echo . && cat /etc/dropbear/authorized_keys && echo .
ssh-rsa AAAlorumipsumrandomgibberish1234567800v99999988888877777766666655555544444443333322221111lorumips+umrandomgibberish1234567800v9999998888887777776666665555554444444333332222111lorumipsum+randomgibberish1234567800v9999998+/andomgibberish1234567800v9999/998888877== me@myhost
ssh-rsa AAAlorumipsumrandomgibberish1234567800v99999988888877777766666655555544444443333322221111lorumips+umrandomgibberish1234567800v9999998888887777776666665555554444444333332222111lorumipsum+randomgibberish1234567800v9999998+/andomgibberish1234567800v9999/998888877== me@myhost
root@ant0:/#
So you can add your rsa key in there if you want, or let Luci do it -- she's helpful that way -- but something tells me I should have read the ~10 or so posts that came after this one (before replying), because you've probably already posted that it works fine now (thanks to jimrome's excellent tag-team tech support action)....in which case, um... Well, in which case the shorter this message, the better!
/me exits stage left
the problem is the persistent cache, is request a complete poweroff of S1 to clear controller cache, or is necessary to flush the cache from console because the modification at LuCI take effect.
