Hello Gordon,
thanks! That seems to be the missing bit.
Is there any kind of documentation available on this topic?
Everything i have read always mentioned not to use idmap at all and delete all
mappings.
Therefore i did the last complete reinstall to have a virgin idmap. Now idmap
shows
no
2012-05-31 6:14, Bob Friesenhahn wrote:
On Tue, 29 May 2012, Jay Heyl wrote:
One idea is to use the external USB drives I've collected to create a
separate pool for the non-critical data. I have four 2TB USB drives that
have been performing flawlessly connected to Windows systems for quite a
We had servers (Solaris 10) setup with external USB drives for ZFS ...
and we had problems.
having more than 8 USB devices attached to a Solaris box caused us
problems, and in some cases having a USB keyboard/mouse connected was
enough to cause the whole USB subsystem to keel over.
after trying
On Thu, May 31, 2012 at 3:54 AM, ths.maila...@yahoo.com wrote:
Hello Gordon,
thanks! That seems to be the missing bit.
Is there any kind of documentation available on this topic?
Well, the idmap man page describes how local UIDs are mapped to SIDs.
It's a fixed, bi-directional algorithm.
A quick update for those who might be following this thread, I started to
collect zilstats, and what I have found is that about once every four days a
transaction takes over half an hour:
TIMEtxgN-Bytes N-Bytes/s N-Max-RateB-Bytes
B-Bytes/s B-Max-Rateops
On Wed, May 30, 2012 at 7:14 PM, Bob Friesenhahn
bfrie...@simple.dallas.tx.us wrote:
I have been using USB drives in a mirror configuration for quite a few
years with zfs. No problems have been encountered due to using zfs. The
main thing I learned is to always export the pool before
On 05/31/12 14:08, Jay Heyl wrote:
That's excellent information. I'm sure you've saved me some future grief.
Thanks.
I have a slightly different story.
I had a Sun X4600 (x64) server running OpenSolaris. I was converting
over from an older system and tried to mirror an external USB drive onto
