2014-07-25 15:41, Larry Rosenman wrote:
On Fri, Jul 25, 2014 at 11:58:48AM +0200, Mark Martinec wrote:
Don't know, I'd guess some network-related memory limit is being hit
on the sending site.

Why not try to decouple the 'zfs send' from a network copy and ssh:
Login to a remote side, do a 'zfs send' to some local temporary file
there, then feed that file to ssh and send it over to a home host,
where it can be piped into some simple program like md5 or 'wc -c'
instead of a zfs recv.

I think I've done that, but it sort of defeats the purpose,

Does it? If you are lucky it would fail again, but this time
you'd know if it was a zfs or a network problem.

as the stream becomes a big file on disk.

If it's an incremental stream and you have sufficient free space
available, than it's all fine and you should try that. If storage
is a problem you may try piping /dev/zero or /dev/random through
ssh to feed a remote consumer, simulating a huge stream, and hope
that it will eventually fail.

I'd like to get some help chasing what parameter(s) need to be fixed here.
Can I get some ideas on:
1) what MIGHT need tweaking
2) would (k|d)trace help?
3) what else could we find out what's being told no memory?

Sorry, not deep enough into fs or network to be able to tell.

My guess it that it's a networking issue (not zfs), and not
directly related to virtual memory or swap or ARC management.

To: freebsd-current@freebsd.org, Freebsd fs <freebsd...@freebsd.org>

If you'd be able to demonstrate that this is unrelated to zfs,
then the freebsd-...@freebsd.org mailing list would perhaps be
a suitable place to continue this thread.

freebsd-current@freebsd.org mailing list
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Reply via email to