AF:
NF:0
PS:10
SRH:1
SFN:
DSR:
MID:<20090416134746.09d80f9a@ripper.onstor.net>
CFG:
PT:0
S:andy.sharp@onstor.com
RQ:
SSV:mail.onstor.net
NSV:
SSH:
R:<dmitry@nexenta.com>
MAID:1
X-Sylpheed-Privacy-System:
X-Sylpheed-Sign:0
SCF:#mh/Mailbox/sent
RMID:#imap/andys@onstor.net@exch1.onstor.net/INBOX	0	49E79810.7090401@nexenta.com
X-Sylpheed-End-Special-Headers: 1
Date: Thu, 16 Apr 2009 13:47:51 -0700
From: Andrew Sharp <andy.sharp@onstor.com>
To: Dmitry Yusupov <dmitry@nexenta.com>
Subject: Re: storage JBOD fault reporting
Message-ID: <20090416134751.23144082@ripper.onstor.net>
In-Reply-To: <49E79810.7090401@nexenta.com>
References: <20090414120057.1c9dd707@ripper.onstor.net>
	<49E51ABD.7040504@nexenta.com>
	<102AB4F33EBBDB4C91915B145C8E9FB31284F9B535@exch1.onstor.net>
	<49E66CE7.1000802@nexenta.com>
	<6041AEBD8A3E264785ABDA2F60B52DB4FEE6E01D@exch1.onstor.net>
	<102AB4F33EBBDB4C91915B145C8E9FB31284F9B542@exch1.onstor.net>
	<49E674C2.4030204@nexenta.com>
	<102AB4F33EBBDB4C91915B145C8E9FB31284F9B548@exch1.onstor.net>
	<49E79810.7090401@nexenta.com>
Organization: Onstor
X-Mailer: Sylpheed-Claws 2.6.0 (GTK+ 2.8.20; x86_64-pc-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

On Thu, 16 Apr 2009 13:41:52 -0700 Dmitry Yusupov <dmitry@nexenta.com>
wrote:

> Brian,
> 
> I really think OnStor can do value add here by implementing SES driver 
> for your appliance. Take a look on recent SES/OpenSolaris putback.

Dmitry,

Can you tell me what that means?  Where/how/what is a putback?  What's
the right way to get it and look at it?  Sorry for being so behind the
curve on things...

I've some documentation on pulling various things from nexenta to
various places at onstor, so if it's a part of that process, just let me
know.

Cheers,

a


> Thanks!
> 
> Brian Stark wrote:
> > Alex,
> > 
> > OK, so I guess we can't rely on FMA to report events relating to power supplies and fans in the server or JBOD.
> > 
> > How have some of your other customers solved this issue?  I'm assuming this topic has come up from some of your other customers, especially since these notifications are a given in the enterprise world.
> > 
> > 
> > Brian
> > 
> > 
> > 
> > -----Original Message-----
> > From: Alex Aizman [mailto:alex@nexenta.com] 
> > Sent: Wednesday, April 15, 2009 4:59 PM
> > To: Brian Stark
> > Cc: 'Dmitry Yusupov'; Andy Sharp
> > Subject: Re: storage JBOD fault reporting
> > 
> > None. The whole subject of SES integration is at its initial stage. We 
> > just started with slot mapping and drive blinking on demand.
> > 
> > Thanks
> > 
> > Brian Stark wrote:
> >> Alex,
> >>
> >> Which of the runners would flag events like power supply and fan failures?  
> >>
> >>
> >> Brian
> >>
> >>
> >> -----Original Message-----
> >> From: Brian Stark 
> >> Sent: Wednesday, April 15, 2009 4:31 PM
> >> To: Alex Aizman
> >> Cc: Dmitry Yusupov; Andy Sharp
> >> Subject: RE: storage JBOD fault reporting
> >>
> >> Alex,
> >>
> >> We're not seeing any events reported by FMA when we pull a power supply out, induce a fan failure, etc. in the server or the JBOD.  When we induce one of these events, 'fmdump' is empty.
> >>
> >>
> >> Brian
> >>
> >>
> >> -----Original Message-----
> >> From: Alex Aizman [mailto:alex@nexenta.com] 
> >> Sent: Wednesday, April 15, 2009 4:25 PM
> >> To: Brian Stark
> >> Cc: Dmitry Yusupov; Andy Sharp
> >> Subject: Re: storage JBOD fault reporting
> >>
> >> It is configured. If we are not getting events, or getting them 
> >> belatedly, that'd be a bug..
> >>
> >> Brian Stark wrote:
> >>> Dmitry,
> >>>
> >>> We've done some reading on FMA, but it's not clear how we configure
> >>> FMA to capture events.  Unlike processor and memory faults, it's also
> >>> not clear that FMA can capture the events like fan and power supply
> >>> failures on either the server or JBOD.
> >>>
> >>> Could you provide some further details on how FMA could be used for
> >>> this?  Do you have examples of how we would configure FMA?
> >>>
> >>> We've also seen some examples of how to use ipmitool to query BMC
> >>> modules in the server.  This would cover the hardware events like fan
> >>> and power supply failures.  However, ipmitool does not appear to be
> >>> installed.
> >>>
> >>>
> >>> Thanks, Brian
> >>>
> >>>
> >>> -----Original Message----- From: Dmitry Yusupov
> >>> [mailto:dmitry@nexenta.com] Sent: Tuesday, April 14, 2009 4:23 PM To:
> >>> Andy Sharp Cc: Alex Aizman; Brian Stark Subject: Re: storage JBOD
> >>> fault reporting
> >>>
> >>> Hi Andrew,
> >>>
> >>> I would let ZFS handle error. We have full blown FMA integration, so
> >>> in case of ZFS detecting error, the e-mail with FMA details will be
> >>> sent. Customer also could dump FMA events using fmadm/fmdump
> >>> commands.
> >>>
> >>> Thanks!
> >>>
> >>> Andrew Sharp wrote:
> >>>> Hi Dmitry,
> >>>>
> >>>> I'm temporarily filling in for Larry as best I can, and I have a 
> >>>> question regarding snmp that I hope you can address.
> >>>>
> >>>> I need to connect up log messages coming in from syslog that are
> >>>> caused by non-disk fault events on the JBOD, like fan or power
> >>>> supply death and so forth.
> >>>>
> >>>> So I need to get those events into either the nexenta fault
> >>>> reporting infrastructure, or simply comb that log data for the
> >>>> particular messages we're looking for and use the snmptrap command
> >>>> to issue an snmp trap.  The second option would be the easiest and
> >>>> quickest from my point of view.  The problem is that it appears
> >>>> that only the snmpd and snmptrapd are built and installed.  I
> >>>> assume that all the snmp commands come from the same source
> >>>> package, so I'm hoping that you can just add those other commands
> >>>> to the distro.
> >>>>
> >>>> Let me know what you think.
> >>>>
> >>>> Thanks,
> >>>>
> >>>> a
> >>>>
> >>>>
> >>
> >>
> > 
> > 
> > 
