Re: [zones-discuss] zone stuck in shutting_down state

2007-08-20 Thread Enda O'Connor ( Sun Micro Systems Ireland) 
Enda O'Connor ( Sun Micro Systems Ireland) wrote:
> Steve Zinck wrote:
>> For future reference, here's an update to my zone hung problem. It 
>> reoccurred late last week so I updated my case with Sun with a crash 
>> dump.  After analysing it, they reported the following:
>>
>> "It appears you are runnning into bug 6570859 "zfs deadlock upon 
>> shutdown" which got closed as a duplicate
>> of 6569877, there is no offical patch yet. There is an IDR."
>>
>> Apparently the patch is part of U4, but since these zones live on ZFS 
>> and afaik U4 isn't going to support upgrading zones on ZFS, I'll 
>> likely apply the IDR. Perhaps move the zones to UFS.  I'll have to 
>> think about it.
>>
> If there is no patch, then it's not in u4, and as there is no patch ( I 
> checked ), then it will be u5 at the earliest I assume.
> 
> BTW you do not have to upgrade to get bug fixes, the official patch will 
> apply fine to your system once it releases ( apply latest patch add 
> first ).
> 
> But it appears the IDR is your only solution for now, post applying the 
> IDR, you will not be able to patch SUNWzfskr package, which means you 
> will not be able to apply the U4 KU 120011/120012, without first running 
> patchrm on the IDR, as IDR's lock the package to stop further patching 
> of the package.
> 
> Just so you are aware.

Ok, so George Wilson clarified things a bit more for me on this one.
The real fix for this is 6447701, which is in u4, the patch will release 
approx mid September ( post official u4 release ).

The patchid's to fiox above will be:
120011-14 ( or higher )  SPARC
120012-14  ( or higher )  X86

These are the u4 KU's and require a lot of patches ( SPARC has 18 
requirements, x86 about the same )

Sorry for any confusion.

Enda

> 
> Enda
>> Thanks,
>>
>> Steve
>>
>> On 7/20/07, *Jeff Victor* <[EMAIL PROTECTED] 
>> > wrote:
>>
>> Hi Steve,
>>
>> When you place the call with Sun Service, ask them to compare your
>> situation to:
>>
>> CR 6547153: GZ should be able to 'halt' a non-global zone (when a
>> NGZ process
>> is in biowait)
>>
>> Steve Zinck wrote:
>>  > I've run into the problem of a zone refusing to cleanly shut down
>> a few
>>  > times now.  It gets stuck in the shutting_down state:
>>  >
>>  >6 as39 shutting_down  /zones/as39
>>  >
>>  >
>>  > Here's ps -efZ |grep as39 from the global zone:
>>  >
>>  > as39root   584 1   0   Jun 24 ?   0:00 zsched
>>  >   globalroot   575 1   0   Jun 24 ?   0:00
>> zoneadmd -z as39
>>  > as39  svctag  4914  4890   0- ?   0:00 
>> 
>>  > as39root  4890 1   0   Jul 04 ?   1:25
>>  > /apps/oracle/oas/10.1.3.1/Apache/Apache/bin/httpd -d
>>  > /apps/oracle/oas/10.1.3.1/
>>  >
>>  > I tried to kill -9 4914 and 4890 with no luck.  I tried to preap
>> (w/ and
>>  > w/o -F) 4914 with no luck.  I've googled and I've seen this
>> discussed in
>>  > the past but I want to know if there's any new suggestions.  I'm
>> going
>>  > to open a call with Sun, what information should I gather for 
>> them?
>>  > halt -d?
>>  >
>>  > This server is a T2000, S10 11/06:
>>  > SunOS gc16 5.10 Generic_118833-36 sun4v sparc SUNW,Sun-Fire-T200
>>  >
>>  > This zone is a whole root zone and it lives on a zfs filesystem.
>>  >
>>  > Thanks,
>>  >
>>  > Steve
>>  >
>>  >
>>  >
>> 
>> 
>>  >
>>  > ___
>>  > zones-discuss mailing list
>>  > zones-discuss@opensolaris.org 
>> 
>>
>> --
>> 
>> -- 
>>
>>
>> Jeff VICTOR  Sun Microsystemsjeff.victor @
>> sun.com 
>> OS AmbassadorSr. Technical Specialist
>> Solaris 10 Zones FAQ:
>> http://www.opensolaris.org/os/community/zones/faq
>> 
>> -- 
>>
>>
>>
>>
>> 
>>
>> ___
>> zones-discuss mailing list
>> zones-discuss@opensolaris.org
> 
> 

___
zones-discuss mailing list
zones-discuss@opensolaris.org

Re: [zones-discuss] zone stuck in shutting_down state

2007-08-20 Thread Enda O'Connor ( Sun Micro Systems Ireland) 
Steve Zinck wrote:
> For future reference, here's an update to my zone hung problem. It 
> reoccurred late last week so I updated my case with Sun with a crash 
> dump.  After analysing it, they reported the following:
> 
> "It appears you are runnning into bug 6570859 "zfs deadlock upon 
> shutdown" which got closed as a duplicate
> of 6569877, there is no offical patch yet. There is an IDR."
> 
> Apparently the patch is part of U4, but since these zones live on ZFS 
> and afaik U4 isn't going to support upgrading zones on ZFS, I'll likely 
> apply the IDR. Perhaps move the zones to UFS.  I'll have to think about it.
> 
If there is no patch, then it's not in u4, and as there is no patch ( I 
checked ), then it will be u5 at the earliest I assume.

BTW you do not have to upgrade to get bug fixes, the official patch will 
apply fine to your system once it releases ( apply latest patch add first ).

But it appears the IDR is your only solution for now, post applying the 
IDR, you will not be able to patch SUNWzfskr package, which means you 
will not be able to apply the U4 KU 120011/120012, without first running 
patchrm on the IDR, as IDR's lock the package to stop further patching 
of the package.

Just so you are aware.

Enda
> Thanks,
> 
> Steve
> 
> On 7/20/07, *Jeff Victor* <[EMAIL PROTECTED] 
> > wrote:
> 
> Hi Steve,
> 
> When you place the call with Sun Service, ask them to compare your
> situation to:
> 
> CR 6547153: GZ should be able to 'halt' a non-global zone (when a
> NGZ process
> is in biowait)
> 
> Steve Zinck wrote:
>  > I've run into the problem of a zone refusing to cleanly shut down
> a few
>  > times now.  It gets stuck in the shutting_down state:
>  >
>  >6 as39 shutting_down  /zones/as39
>  >
>  >
>  > Here's ps -efZ |grep as39 from the global zone:
>  >
>  > as39root   584 1   0   Jun 24 ?   0:00 zsched
>  >   globalroot   575 1   0   Jun 24 ?   0:00
> zoneadmd -z as39
>  > as39  svctag  4914  4890   0- ?   0:00 
>  > as39root  4890 1   0   Jul 04 ?   1:25
>  > /apps/oracle/oas/10.1.3.1/Apache/Apache/bin/httpd -d
>  > /apps/oracle/oas/10.1.3.1/
>  >
>  > I tried to kill -9 4914 and 4890 with no luck.  I tried to preap
> (w/ and
>  > w/o -F) 4914 with no luck.  I've googled and I've seen this
> discussed in
>  > the past but I want to know if there's any new suggestions.  I'm
> going
>  > to open a call with Sun, what information should I gather for them?
>  > halt -d?
>  >
>  > This server is a T2000, S10 11/06:
>  > SunOS gc16 5.10 Generic_118833-36 sun4v sparc SUNW,Sun-Fire-T200
>  >
>  > This zone is a whole root zone and it lives on a zfs filesystem.
>  >
>  > Thanks,
>  >
>  > Steve
>  >
>  >
>  >
> 
>  >
>  > ___
>  > zones-discuss mailing list
>  > zones-discuss@opensolaris.org 
> 
> --
> --
> 
> Jeff VICTOR  Sun Microsystemsjeff.victor @
> sun.com 
> OS AmbassadorSr. Technical Specialist
> Solaris 10 Zones FAQ:
> http://www.opensolaris.org/os/community/zones/faq
> --
> 
> 
> 
> 
> 
> ___
> zones-discuss mailing list
> zones-discuss@opensolaris.org

___
zones-discuss mailing list
zones-discuss@opensolaris.org

Re: [zones-discuss] zone stuck in shutting_down state

2007-08-19 Thread Steve Zinck 
For future reference, here's an update to my zone hung problem. It
reoccurred late last week so I updated my case with Sun with a crash dump.
After analysing it, they reported the following:

"It appears you are runnning into bug 6570859 "zfs deadlock upon shutdown"
which got closed as a duplicate
of 6569877, there is no offical patch yet. There is an IDR."

Apparently the patch is part of U4, but since these zones live on ZFS and
afaik U4 isn't going to support upgrading zones on ZFS, I'll likely apply
the IDR. Perhaps move the zones to UFS.  I'll have to think about it.

Thanks,

Steve

On 7/20/07, Jeff Victor <[EMAIL PROTECTED]> wrote:
>
> Hi Steve,
>
> When you place the call with Sun Service, ask them to compare your
> situation to:
>
> CR 6547153: GZ should be able to 'halt' a non-global zone (when a NGZ
> process
> is in biowait)
>
> Steve Zinck wrote:
> > I've run into the problem of a zone refusing to cleanly shut down a few
> > times now.  It gets stuck in the shutting_down state:
> >
> >6 as39 shutting_down  /zones/as39
> >
> >
> > Here's ps -efZ |grep as39 from the global zone:
> >
> > as39root   584 1   0   Jun 24 ?   0:00 zsched
> >   globalroot   575 1   0   Jun 24 ?   0:00 zoneadmd -z
> as39
> > as39  svctag  4914  4890   0- ?   0:00 
> > as39root  4890 1   0   Jul 04 ?   1:25
> > /apps/oracle/oas/10.1.3.1/Apache/Apache/bin/httpd -d
> > /apps/oracle/oas/10.1.3.1/
> >
> > I tried to kill -9 4914 and 4890 with no luck.  I tried to preap (w/ and
> > w/o -F) 4914 with no luck.  I've googled and I've seen this discussed in
> > the past but I want to know if there's any new suggestions.  I'm going
> > to open a call with Sun, what information should I gather for them?
> > halt -d?
> >
> > This server is a T2000, S10 11/06:
> > SunOS gc16 5.10 Generic_118833-36 sun4v sparc SUNW,Sun-Fire-T200
> >
> > This zone is a whole root zone and it lives on a zfs filesystem.
> >
> > Thanks,
> >
> > Steve
> >
> >
> > 
> >
> > ___
> > zones-discuss mailing list
> > zones-discuss@opensolaris.org
>
> --
> --
> Jeff VICTOR  Sun Microsystemsjeff.victor @ sun.com
> OS AmbassadorSr. Technical Specialist
> Solaris 10 Zones FAQ:http://www.opensolaris.org/os/community/zones/faq
> --
>
___
zones-discuss mailing list
zones-discuss@opensolaris.org

Re: [zones-discuss] zone stuck in shutting_down state

2007-07-20 Thread Jerry Jelinek 
Steve Zinck wrote:
> I'm going to open a
> call with Sun, what information should I gather for them?  halt -d?

Yes, a system crash dump of the system in this state is what is needed.

Thanks,
Jerry
___
zones-discuss mailing list
zones-discuss@opensolaris.org

Re: [zones-discuss] zone stuck in shutting_down state

2007-07-20 Thread Jeff Victor 
Hi Steve,

When you place the call with Sun Service, ask them to compare your situation to:

CR 6547153: GZ should be able to 'halt' a non-global zone (when a NGZ process 
is in biowait)

Steve Zinck wrote:
> I've run into the problem of a zone refusing to cleanly shut down a few 
> times now.  It gets stuck in the shutting_down state:
> 
>6 as39 shutting_down  /zones/as39
> 
> 
> Here's ps -efZ |grep as39 from the global zone:
> 
> as39root   584 1   0   Jun 24 ?   0:00 zsched
>   globalroot   575 1   0   Jun 24 ?   0:00 zoneadmd -z as39
> as39  svctag  4914  4890   0- ?   0:00 
> as39root  4890 1   0   Jul 04 ?   1:25 
> /apps/oracle/oas/10.1.3.1/Apache/Apache/bin/httpd -d 
> /apps/oracle/oas/10.1.3.1/
> 
> I tried to kill -9 4914 and 4890 with no luck.  I tried to preap (w/ and 
> w/o -F) 4914 with no luck.  I've googled and I've seen this discussed in 
> the past but I want to know if there's any new suggestions.  I'm going 
> to open a call with Sun, what information should I gather for them?  
> halt -d?
> 
> This server is a T2000, S10 11/06:
> SunOS gc16 5.10 Generic_118833-36 sun4v sparc SUNW,Sun-Fire-T200
> 
> This zone is a whole root zone and it lives on a zfs filesystem.
> 
> Thanks,
> 
> Steve
> 
> 
> 
> 
> ___
> zones-discuss mailing list
> zones-discuss@opensolaris.org

-- 
--
Jeff VICTOR  Sun Microsystemsjeff.victor @ sun.com
OS AmbassadorSr. Technical Specialist
Solaris 10 Zones FAQ:http://www.opensolaris.org/os/community/zones/faq
--
___
zones-discuss mailing list
zones-discuss@opensolaris.org