Quote from: rjk on March 02, 2012, 04:45:00 AM

Quote from: smickles on March 02, 2012, 04:42:35 AM

Quote from: btc_artist on March 02, 2012, 04:41:13 AM

Quote from: smickles on March 02, 2012, 04:38:02 AM

Quote from: bitcoinBull on March 02, 2012, 04:34:11 AM

Quote from: bearbones on March 02, 2012, 04:26:43 AM

Quote from: cablepair on March 02, 2012, 04:23:08 AM

i mean seriously, could not this whole thing been prevented if the wallet was just encrypted?

Obviously the software running against the hot wallet has to have access to it. This means that if someone roots the server, they'll be able to have the same access to the hot wallet. Encryption would not have entered into it.

Zhou, good on you for covering this! I'm having a hard enough time covering the BTCinch theft; I can only imagine how pissed you are at linode.

In this case, encryption would have protected the wallet because the attacker was only able to get root access after a reboot.

why would a reboot stop the attacker from seeing the wallet being unencrypted during the next use?

You have to enter the wallet password/passphrase after rebooting/restarting bitcoin.

am i missing something here? wouldn't that entry be exactly what the attacker would be waiting for?

Pretty sure such a random suspicious reboot would cause the poolop to review the server before entering any creds anywhere. Especially when his Linode access manager says that there was a login to his account a few minutes before, not caused by him.

+1, the idea that this hacker is sitting here watching a packet sniffer or a keylogger and the admin of the server with an encrypted wallet holding $200k+ is not going to think something suspicions is preposterous

it would take multiple fails for this scenario to be successful and the bottom line is an encrypted wallet would likely have saved this money. The problem is these web applications have not been developed to the level where they are able to interact with encrypted wallets. point blank.