Re: GELI encryption and HDD critical temperature
Hi, this problem is caused by how the drives are installed in the machine. Can you add a fan? Erich s.g. wrote: Guys, I notice occasional overheating of my GELI-encrypted hard drives followed by the reboot. This happens when there is heavy activity on the drive - eg when trying to dump partitions. There are 4 drives i have, three encrypted and fourth is plaintext. According to smartctl -a, the temperature of the encrypted drives is ~59C. The temperature of the unencrypted drive is, however, ~41C, according to the same smartctl -a. All the four drives in question are Seagate ST3400620AS/3.AAC, GELI encryption is default AES, no data authentication. This is on 6.2-RELEASE-p8. Am I right assuming GELI encryption is the reason for such a "global warming"? Thanks in advance, Grigorian ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]" ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: GELI encryption and HDD critical temperature
On Tue, Feb 19, 2008 at 11:41:18AM +0300, s.g. wrote: > According to smartctl -a, the temperature of the encrypted drives is ~59C. > The temperature of the unencrypted drive is, however, ~41C, according to > the same smartctl -a. The CPU has to work extra hard to encrypt/decrypt, and it is possible that the extra heat this generates is absorbed by the HDDs. But since the only drive that's overheating is the encrypted one, it seems to be something else. > Am I right assuming GELI encryption is the reason for such a "global > warming"? I don't know. But I've noticed that when drives access GBDE-encrypted partitions (I didn't try with GELI yet), they are much louder (head seeking). It seems they seek more often on encrypted than non-encrypted partitions. Perhaps caching is turned off at some point up the chain? If that's the case, it is no wonder that encrypted partitions tend to result in higher drive temps (and faster drive wear). Is there a way to measure the number of head seeks in near real-time to confirm or disprove this? > Grigorian -cpghost. -- Cordula's Web. http://www.cordula.ws/ ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
RE: GELI encryption and HDD critical temperature
> -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of cpghost > Sent: Thursday, 21 February 2008 11:42 p.m. > To: s.g. > Cc: freebsd-questions@freebsd.org > Subject: Re: GELI encryption and HDD critical temperature > > On Tue, Feb 19, 2008 at 11:41:18AM +0300, s.g. wrote: > > According to smartctl -a, the temperature of the encrypted > drives is ~59C. > > The temperature of the unencrypted drive is, however, > ~41C, according to > > the same smartctl -a. > > I don't know. But I've noticed that when drives access GBDE-encrypted > partitions (I didn't try with GELI yet), they are much louder (head > seeking). It seems they seek more often on encrypted than > non-encrypted > partitions. Perhaps caching is turned off at some point up the chain? > If that's the case, it is no wonder that encrypted partitions tend to > result in higher drive temps (and faster drive wear). It was explained by another poster, I don't remember when or by whom, that GBDE writes sectors to disk in a pseudorandom fashion to make cryptanalysis more difficult. This would explain the seeking/noise on a GBDE disk. A question I have which is related to all of this: Does GELI write sectors in this pseudorandom fashion as well? And, if so is there a way to turn this off so that things are written contiguously? This could be useful for those wishing to encrypt things for most "normal" threats, such as your teenage neighbour breaking into your house and stealing your bitchin computer, while minimising the performance hit of pseudorandom sector writes. Cheers, Brent ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"