"Hurrah! 2.0 has been released!" said Matthew Dillon, announcing the eighth major release of DragonFly BSD. This release is the first to include HAMMER, a new clustering filesystem that already boasts an impressive list of features, including: "crash recovery on-mount, no fsck; fine-grained snapshots, snapshot management, snapshot-support for filesystem-wide data integrity checks; historically accessible by default; mirroring: queueless incremental mirroring, master to multi-slave; undo and rollback; reblocking; multi-volume, maximum storage capacity of 1-Exabyte." Other highlighted changes in this release include, "native fairq-queue implementation using ALTQ, for PF", and "native connection state recovery to PF, so router reboots do not drop active TCP connections."
The latest version of DragonFly BSD can be downloaded from a mirror. The download page explains:
"DragonFly CDs are 'live', which means that the CD will boot your system and let you log in as root (no password). You can use this feature to check for hardware compatibility and play with DragonFly a little before actually installing it on your hard drive."
From: Matthew Dillon <dillon@...>
Subject: HAMMER Update - 16-July-2008
Date: Jul 17, 2:31 am 2008
HAMMER is looking nice and stable for the release. I've moved on to
working through its error handling, getting rid of Debugger() calls
and assertions in the I/O error path, and adjusting the code so a
write error forces the mount into read-only mode.
It's kinda fun. I'm testing it by taking an external USB2 hard drive
and mounting it w/ HAMMER, then unplugging it in the middle of a big
cpdup. I've found a couple of panics in CAM related to the device
going away unexpectedly, and stuff like that, which I'm fixing as I
go.
I expect to get it working nicely by thursday afternoon and will commit
and MFC it for 2.0 then. Hopefully I'll get it to the point where
I can pop the USB connector in and out without crashing the machine
or corrupting the HAMMER filesystem. Mind you, once you pull the drive
you have to umount -f / remount HAMMER and will lose unflushed file
data, but the filesystem itself shouldn't become corrupt. I'm not
pulling the disk's power, just its USB connector :-)
-Matt
Matthew Dillon
<dillon@backplane.com>From: Matthew Dillon <dillon@...>
Subject: 2.0 BRANCH FREEZE - TODAY IS THE LAST DAY FOR COMMITS TO 2.0!
Date: Jul 18, 4:10 pm 2008
Today (Friday) is the last day people can commit to the 2.0 branch.
I would appreciate it if a couple of people could do basic testing of
release builds for 2.0 today (cd /usr/src/nrelease; make installer release).
I will be building and testing the final release ISOs Saturday morning
and letting them propagate to our mirrors overnight.
The release will be officially announced on Sunday!
-Matt
Matthew Dillon
<dillon@backplane.com>From: Matthew Dillon <dillon@...>
Subject: HEADS UP - last minute (non-bug-fix) HAMMER change to master/slave operation
Date: Jul 19, 1:58 pm 2008
I've decided to remove some confusion from the HAMMER utility, and also
remove some of the master id code from the filesystem, which was
designed for multi-master mode but really added confusion to the
whole master-slave setup since we don't support multi-master mode
in this release.
PFS's will now only be master or slave, there will be no master id
(0-15) setting and there will not be a no-mirror mode. A single
whole-mount master id or no-mirror mode can be specified at mount
time which is how I had it originally. That way the master_id is
out of sight and out of mind for the release.
I am also removing the 'master' and 'slave' options to pfs-update,
simplifying the PFS features. Now you can create a master, create a
slave, upgrade a slave to a master, or downgrade a master to a slave.
The confusing 'no-mirror', 'master=' and 'slave' options have been
removed.
I'll update the hammer.5 and hammer.8 docs appropriately, as well.
-MattFrom: Matthew Dillon <dillon@...>
Subject: Mirrors don't have ISOs yet - waiting on release announcement
Date: Jul 20, 1:33 pm 2008
The mirrors don't have the ISOs yet. It looks like we will have to
wait on the 2.0 announcement for another day. I am going to put up a
filler page today.
People with administrative control over their mirror sites, please
try to kick the mirrors in the shin so they continue attempting to
get the ISO from crater's iso-images directory. My outgoing bandwidth
is maxed out, but well managed. I can get 5KBytes/sec testing from
an outside host so I'm not sure why the mirrors stalled out overnight.
In anycase, we're there, we're just not ALL there :-)
-Matt
Matthew Dillon
<dillon@backplane.com>From: Matthew Dillon <dillon@...>
Subject: DragonFly 2.0 - RELEASED!
Date: Jul 21, 12:50 pm 2008
Hurrah! 2.0 has been released!
I am continuing to finish up the release document and will also be
adding a new section on the HAMMER filesystem. Three mirrors still
don't have the ISO and I have temporarily commented them out. We'll
run with what we have.
-Matt
Impressive
Impressive list of features!
How does it work in real life? (No, I'm not being a bastard, I'm curious)
what do you mean by "real life"?
what do you mean by "real life"? performance statistics, stability?
More or less an overall
More or less an overall evaluation, but performance and stability is probably the most important parts, yeah.
The only thing I remember
The only thing I remember seeing was this
http://people.freebsd.org/~kris/scaling/dfly.html
a lot has changed since these results were taken, but clearly there is still a way to go.
All the same DragonflyBSD is an exciting project and is well worth a look.
Time goes backward
Is it intended that the emails are given in reverse order? And if so, why?