On Jan 4, 2009, at 5:12 PM, Alan Bell wrote:
Geir Magnusson Jr. wrote:
On Jan 4, 2009, at 2:59 PM, Damien Katz wrote:
What you say is possible, but doesn't seem too likely as I don't
ever recall any report of a corrupt db on Linux, but before the
F_FILESYNC change in Erlang we would see reports of corruptions on
OS X due to power loss. Since the change, we've had no reports of
corrupt databases on OS X or Linux.
Have you seen the correlation between the # of pirates and global
warming? :)
There may be correlation here between the Erlang update and
corruption, but I don't see the causality - fsync() on linux
should only get as far as the on-device buffering if it does write
buffering, whereas if the OS X docs are truthful, the F_FILESYNC
fcntl() pushs the bits to the rust on the platter. Maybe they
fixed something else in that erlang update, or maybe the universe
is smiling upon CouchDB users or ....
. . . maybe the ec2 virtual machine instance and abstracted storage
thingie isn't getting to the rust, even though from the point of
view of the inside of the VM it thinks it has. I don't know for
sure, but a comparitive on a real Linux box would be worthwhile. Do
you have some simple benchmark code to share and I will give it a go
on a few Linux machines. Would also be interesting to compare SSD to
rotating HDD.
Oh, ec2 was way faster, but I have no clue where the storage really is
- I was going to re-install on the /mnt partition,
I'll put the code on a github project later - was just thinking about
doing exactly that. Thanks - I appreciate hearing how your tests go.
geir
