On 03/05/2016 19:31, Garrett D'Amore wrote:
> 1. Will sysevents in the core create challenges for portability to other OS?
> (Matt? Andriy? Brian?)
It seems like in FreeBSD we already have a compatibility shim for sysevents, at
least log_sysevent.
--
Andriy Gapon
-
I'd be happy with just a cmn_err for now. You could leave a comment behind
indicating a plan to fix with a kstat later.
Sent from my iPhone
> On May 3, 2016, at 4:15 PM, Jerry Jelinek wrote:
>
> At Joyent we have a bug open to add a kstat to track sysevent behavior
> (https://smartos.org/bug
Garrett D'Amore wrote:
A few thoughts here.
1. Will sysevents in the core create challenges for portability to other
OS? (Matt? Andriy? Brian?)
Ohoh and us!
Not entirely sure what it is, it looks to be a kernel event system,
where you will send out the spa_history events, but there are
At Joyent we have a bug open to add a kstat to track sysevent behavior (
https://smartos.org/bugview/OS-5112). I already have an initial
implementation of this, but during our preliminary internal review I was
asked to track more events and add more values. Right now this is on the
back-burner, but
On May 3, 2016, at 5:32 PM, Garrett D'Amore wrote:
> Oh, while here, shouldn’t we syslog on sysevent failures?
"Yes please" I said with deep longing in my voice..
---
openzfs-developer
Archives: https://www.listbox.com/member/archive/274414/=now
RSS Feed:
Oh, while here, shouldn’t we syslog on sysevent failures? Or at least
figure out how to bump a stat or something? The fact that we failed to
allocate or post a sysevent is not recorded *anywhere*, and that seems
pretty unfortunate.
On Tue, May 3, 2016 at 9:31 AM, Garrett D'Amore wrote:
> A few
A few thoughts here.
1. Will sysevents in the core create challenges for portability to other
OS? (Matt? Andriy? Brian?)
2. Can we please use ILLUMOS_KERN_PUB instead of SUNW_KERN_PUB for these
events? (spa_history.c line 212)
Otherwise it seems reasonable to me.
On Thu, Apr 28, 2016 at 8:0
