* Markus Hennecke <markus-henne...@markus-hennecke.de> [2009-04-17 08:06]: > Marco Peereboom schrieb: >> I work with people that run io tools against flash parts. I still have >> to see it fail too. Your puny little firewall will never write more to >> it than a month long stress test. This write fatigue argument is very >> silly. > > Generalization is always false. > I killed a 1GB SanDisk CF Card because of excessive logging of OpenVPN > Connections from WLAN Clients which unfortunately had power saving > enabled and dropped the connection every few minutes. Took me around 2 > or 3 weeks, I just forgot to reduce the log level. Perhaps those stress > tests are not stressing enough? That Card was a little bit older, but > seldom used, so there is a good chance that that scenario no longer > applies.
and I bet it didn't die because of the excessive writes but something else. And until you show me proof that all reserve cells were mapped in and in use I cannot be convinced of the opposite. -- Henning Brauer, h...@bsws.de, henn...@openbsd.org BS Web Services, http://bsws.de Full-Service ISP - Secure Hosting, Mail and DNS Services Dedicated Servers, Rootservers, Application Hosting - Hamburg & Amsterdam
