On 07/21/2015 07:56 PM, Scott Dowdle wrote:
Greetings,

----- Original Message -----
ZFS is really "The Last Word in File Systems",
and now you can just use it for free,
without reinventing the wheel.

OpenVZ + ZFS or Virtuozzo + ZFS == atom bomb,
killer feature with horrible devastation power.

Or - you just forcing users to migrate from OpenVZ
to CentOS+KVM over ZFS and/or CentOS+Docker over ZFS.
Whatever.  So many ZFS users seem to be such fanatics they will abandon 
anything that gets in its way... while none of the top 10 Linux distros will 
ship it.  Folks like Jesse Smith from Distrowatch say there is nothing wrong 
with distros shipping ZFS... as long as it continues to be packaged separately 
from the kernel (a module rather than compiled in)... but still... no one ships 
it.  I believe Debian is working on changing that and I wish them luck.

I've tried it.  I've read the recipes.  Some say you have to dedicate 1GB of 
RAM for every TB of storage.  To build a high performance ZFS-based fileserver 
you really want to custom design the thing with the right combination of read 
cache disks, write cache disks, etc.  It has compression, encryption, dedup 
(not sure if that is in the Linux version yet), etc.  I'm guessing if you just 
want to ZFS for local stuff (VMs, containers, server applications, etc) you 
don't have to worry as much getting an optimal setup as you would for a 
dedicated fileserver.

I second that. ZFS seems to be pretty hungry for RAM, and the requirements
are even higher if you like deduplication feature to work. That means either
lower container density (you can run less CTs if you use ploop), or higher memory
requirements (you need more RAM to run same amount of CTs on ZFS).

I haven't really had a reason to use it.  ZFS + OpenVZ = atomic bomb?  Whatever.

I'd prefer to see BTRFS mature... and once that is in every Linux distro by 
default... and widely deployed... I don't think ZFS will be that relevant 
except among the fanatics.  Now having said that I realize it could take years 
before BTRFS is considered good enough by most folks.  I certain hope it 
doesn't take that long but who knows?  No need to tell me how much BTRFS sucks 
and ZFS rocks.

Please don't provide me with why ZFS is the god of filesystems.  I've heard it 
all before.  If you use and like all of those features and ZFS works great for 
you... go for it... more power to you.

Regarding OpenVZ checkpoint / restore and live migration... it has worked well 
for me since it was originally released in 2007 (or was it 2008?).  While I've 
had a few kernel panics in the almost 10 years I've been using OpenVZ (starting 
with the EL4-based kernel)... I can't remember what year I had my last one.

I see people come into the #openvz IRC channel with bugs all of the time.  The 
vast majority of the time it turns out they are way behind the current stable 
versions of the vzkernel and vzctl.  They really do fix bugs in every release 
so why people seem to think it is ok to ignore updates for months or years is 
beyond me.  I have no idea if you do or not... but hopefully you can feel my 
pain.  Are there bugs in the bug reporting system?  Sure.  People say Debian is 
the most stable Linux distro around (I'm not a Debian user) but if you look in 
their but reporting system I'm sure there are thousands (or more likely, tens 
of thousands) of bug reports.  I guess one should expect that with tens of 
thousands of packages... but my point is there will always be bugs... but to 
point at a bug report and give up saying that it isn't stable because of bug 
report x... or that some people have had panics at some point in history... 
well, that isn't very reflective of the overall picture!
. !
   Nothing personal.  We just disagree on a few topics.  We probably agree on 
way more things though.

I don't use checkpoint and restore directly... but do with vzmigrate.  Every 
time I upgrade vzkernel I use live migration... so I can have as much container 
uptime as possible.  I've had a few times when live migration didn't work but 
in every case it failed safely and I was able to do an offline migration 
instead.  The times it didn't work were when my CPUs differed greatly between 
hosts... or the vzkernel I was running differed too much (I run testing kernels 
on a host or two).  If you don't have a use for offline nor online migration... 
ok... but lots of people do use it... and if ZFS means it can't be used... that 
is just another reason (for me) not to use ZFS.

TYL,



_______________________________________________
Users mailing list
Users@openvz.org
https://lists.openvz.org/mailman/listinfo/users

Reply via email to