[zfs-discuss] sending stream too slow on busy system

Ruben Wisniewski ruben at vfn-nrw.de
Fri May 27 02:14:45 EDT 2016


Me too, both of it. ;)

> Am 27.05.2016 um 03:48 schrieb Tren Blackburn via zfs-discuss <zfs-discuss at list.zfsonlinux.org>:
> 
> On May 26, 2016 at 5:15:42 PM, Benjamin Smith via zfs-discuss
> (zfs-discuss at list.zfsonlinux.org(mailto:zfs-discuss at list.zfsonlinux.org))
> wrote:
> 
>>> On Monday, May 23, 2016 11:54:09 AM Paul Dagnelie via zfs-discuss wrote:
>>> you
>>> probably want to look into resumable send/recv; if your send is going to
>>> take a week, you don't want to have to restart on day 6 because your
>>> network dropped some packets on the floor.
>> 
>> I developed a tool called "relentlesspipe" a while that essentially does a
>> retrying MITM during a send/receive to handle this exact situation. You can
>> yank the network cable, even change IP addresses of the sending/receiving host
>> and as long as the DNS names resolve and neither system reboots, it will pick
>> up and resume just fine.
>> 
>> It a bit of set up, but once implemented, our DR and Archive replication has
>> been basically flawless. The only time it's at all painful is when you manually
>> kill it, you have to kill processes manually with `ps aux` and `kill -9`.
>> (which happens perhaps 1x/year for us) It's written in PHP; here's the
>> documentation for it, let me know if you want a copy:
> 
> I would love a copy of this. There have been times I would have killed
> for something like this.
> 
> Thank you for sharing!
> 
> Regards,
> 
> Tren
> _______________________________________________
> zfs-discuss mailing list
> zfs-discuss at list.zfsonlinux.org
> http://list.zfsonlinux.org/cgi-bin/mailman/listinfo/zfs-discuss



More information about the zfs-discuss mailing list