PV Status NOT available

linux-lvm Rupert Hair 3 participants 8 messages Feb 10, 2003 – Feb 13, 2003

Share & tools

Rupert Hair

author

2003-02-10 06:27 UTC #0 lore raw

Hi Everyone.

I have a rather strange 'problem' with my LVM setup. I am using a very standard install of Redhat 8.0 (Kernel 2.4.18-14smp (I guess the "-14smp" is some Redhat version number.)) with LVM version 1.0.3(19/02/2002). I used the Redhat installer to create a single VG (datavg) and LV (datalv) made up of 3 PVs (hdd5 hdd6 hdd7 (Yes all three PVs are on the same disk as this is just a test setup.)). This all went fine and the installer created an ext3 filesystem on top of the LV. I have put some data on the filesystem and rebooted a few times to test it out and everything works great.

Now the 'problem': two of my PVs (hdd5 hdd7) have a status of "NOT available" (reported by pvdisplay). This is very strange as the whole setup seems to work fine and all three PVs are on the same disk (which is working fine). I have attached various files which are the outputs of some of the LVM utilities that seemed useful.

I would be very grateful if anyone could shed any light on this problem and what I have done wrong.

Thanks.
Rupert

Heinz J . Mauelshagen

2003-02-10 07:38 UTC #1 lore raw

Rupert,

do all 3 PVs show up active (the [] following 'PV:' should contain 'AA' unless you set a PV to inallocatable) in /proc/lvm/global ?

if so pvdisplay (pv_status in the library respectively) doesn't seem to get the PV_ACTIVE status from the kernel correctly.

Quoted 22 lines · Rupert Hair

Hi Everyone.
I have a rather strange 'problem' with my LVM setup. I am using a very standard install of Redhat 8.0 (Kernel 2.4.18-14smp (I guess the "-14smp" is some Redhat version number.)) with LVM version 1.0.3(19/02/2002). I used the Redhat installer to create a single VG (datavg) and LV (datalv) made up of 3 PVs (hdd5 hdd6 hdd7 (Yes all three PVs are on the same disk as this is just a test setup.)). This all went fine and the installer created an ext3 filesystem on top of the LV. I have put some data on the filesystem and rebooted a few times to test it out and everything works great.
Now the 'problem': two of my PVs (hdd5 hdd7) have a status of "NOT available" (reported by pvdisplay). This is very strange as the whole setup seems to work fine and all three PVs are on the same disk (which is working fine). I have attached various files which are the outputs of some of the LVM utilities that seemed useful.
I would be very grateful if anyone could shed any light on this problem and what I have done wrong.
Thanks.
Rupert

Signature · 15 lines

Regards,
Heinz    -- The LVM Guy --

*** Software bugs are stupid.
    Nevertheless it needs not so stupid people to solve them ***

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Heinz Mauelshagen                                 Sistina Software Inc.
Senior Consultant/Developer                       Am Sonnenhang 11
                                                  56242 Marienrachdorf
                                                  Germany
[email protected]                           +49 2626 141200
                                                       FAX 924446
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Patrick Boutilier

2003-02-10 08:02 UTC #2 lore raw

I have the same problem on RedHat 8.0

# pvdisplay /dev/sda6

--- Physical volume ---

PV Name               /dev/sda6
VG Name               imap_vg
PV Size               261.11 GB [547591527 secs] / NOT usable 16.25 MB [LVM: 189 KB]
PV#                   1
PV Status             NOT available
Allocatable           yes (but full)
Cur LV                1
PE Size (KByte)       16384
Total PE              16710
Free PE               0
Allocated PE          16710
PV UUID               O7WnwM-5ASt-AsiV-8BlZ-h6KP-NjFl-cdY33C

# cat /proc/lvm/global

LVM driver LVM version 1.0.6(25/10/2002)

Total: 1 VG 1 PV 1 LV (1 LV open 1 times)

Global: 271038 bytes malloced IOP version: 10 63 days 17:26:47 active

VG: imap_vg [1 PV, 1 LV/1 open] PE Size: 16384 KB
Usage [KB/PE]: 273776640 /16710 total 273776640 /16710 used 0 /0 free
PV: [AA] sda6 273776640 /16710 273776640 /16710

        0 /0
    LV:  [AWDL  ] imap_lv                  273776640 /16710    1x open

Quoted 35 lines · Heinz J . Mauelshagen

Rupert,
do all 3 PVs show up active (the [] following 'PV:' should contain 'AA' unless you set a PV to inallocatable) in /proc/lvm/global ?
if so pvdisplay (pv_status in the library respectively) doesn't seem to get the PV_ACTIVE status from the kernel correctly.
Quoted 22 lines · Rupert Hair
Hi Everyone.
I have a rather strange 'problem' with my LVM setup. I am using a very standard install of Redhat 8.0 (Kernel 2.4.18-14smp (I guess the "-14smp" is some Redhat version number.)) with LVM version 1.0.3(19/02/2002). I used the Redhat installer to create a single VG (datavg) and LV (datalv) made up of 3 PVs (hdd5 hdd6 hdd7 (Yes all three PVs are on the same disk as this is just a test setup.)). This all went fine and the installer created an ext3 filesystem on top of the LV. I have put some data on the filesystem and rebooted a few times to test it out and everything works great.
Now the 'problem': two of my PVs (hdd5 hdd7) have a status of "NOT available" (reported by pvdisplay). This is very strange as the whole setup seems to work fine and all three PVs are on the same disk (which is working fine). I have attached various files which are the outputs of some of the LVM utilities that seemed useful.
I would be very grateful if anyone could shed any light on this problem and what I have done wrong.
Thanks.
Rupert

Rupert Hair

author

2003-02-10 09:31 UTC #3 lore raw

In reply to Heinz J . Mauelshagen

Quoted 2 lines

do all 3 PVs show up active (the [] following 'PV:' should contain 'AA' unless you set a PV to inallocatable) in /proc/lvm/global ?

LVM driver LVM version 1.0.5(15/07/2002)

Total: 1 VG 3 PVs 1 LV (1 LV open 1 times)

Global: 179278 bytes malloced IOP version: 10 5 days 1:33:43 active

VG: datavg [3 PV, 1 LV/1 open] PE Size: 32768 KB
Usage [KB/PE]: 357171200 /10900 total 357171200 /10900 used 0 /0 free
PVs: [AA] hdg1 119996416 /3662 119996416 /3662
0 /0

       [AA] hde1                  119996416 /3662    119996416 /3662
0 /0
       [AA] hdf1                  117178368 /3576    117178368 /3576
0 /0
    LV:  [AWDL  ] datalv                   357171200 /10900    1x open

Looks like Patrick and I have the same problem. Thanks for your help Heinz.

Rupert

Heinz J . Mauelshagen

2003-02-11 09:34 UTC #4 lore raw

Quoted 13 lines · Rupert Hair

Quoted 3 lines
do all 3 PVs show up active (the [] following 'PV:' should contain 'AA' unless you set a PV to inallocatable) in /proc/lvm/global ?
LVM driver LVM version 1.0.5(15/07/2002)
Total: 1 VG 3 PVs 1 LV (1 LV open 1 times)
Global: 179278 bytes malloced IOP version: 10 5 days 1:33:43 active
VG: datavg [3 PV, 1 LV/1 open] PE Size: 32768 KB
Usage [KB/PE]: 357171200 /10900 total 357171200 /10900 used 0 /0 free
PVs: [AA] hdg1 119996416 /3662 119996416 /3662
0 /0

Quoted 7 lines

       [AA] hde1                  119996416 /3662    119996416 /3662
0 /0
       [AA] hdf1                  117178368 /3576    117178368 /3576
0 /0
    LV:  [AWDL  ] datalv                   357171200 /10900    1x open

Looks like Patrick and I have the same problem. Thanks for your help Heinz.

Hmmm, pretty od really. pvdisplay calls the pv_status() lib function which retrieves the information from the kernel. That's why i asked for the proc output which shows that the PVs are _active_ and allocatable.

Wait, need to preserve the core flags to make this work...

Could you test if the following patch to pvdisplay.c helps ?

pvdisplay.c 11 Feb 2003 15:31:51 -0000 renamed from pvdisplay.c 6 Feb 2003 15:05:30 -0000 1.25 +3 -1

diff -u -B -u -r1.25 pvdisplay.c
@@ -42,6 +42,7 @@
  *    06/02/2002 - cope with changed pv_get_size() return cast
  *    07/02/2002 - fixes for > 1TB support
  *    17/02/2002 - avoid pv_get_size() altogether
+ *    11/02/2002 - fixed bug preventing correct 'active' status displayed [HM]
  *
  */

@@ -231,7 +232,8 @@
          return LVM_EPVDISPLAY_PV_CHECK_CONSISTENCY;
       }
 
-      /* restore the from disk pointer */
+      /* restore the from disk pointer _but_ preserve kernel status */
+      pv_from_disk->pv_status = pv->pv_status;
       pv = pv_from_disk;
 
       if ( opt_c == 0) {

Quoted 10 lines

Rupert
_______________________________________________
linux-lvm mailing list
[email protected]
http://lists.sistina.com/mailman/listinfo/linux-lvm
read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/

Regards,
Heinz -- The LVM Guy --

*** Software bugs are stupid.
Nevertheless it needs not so stupid people to solve them ***

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Heinz Mauelshagen Sistina Software Inc.
Senior Consultant/Developer Am Sonnenhang 11
56242 Marienrachdorf Germany
[email protected] +49 2626 141200

                                                       FAX 924446

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Patrick Boutilier

2003-02-11 10:44 UTC #5 lore raw

Works for me.

Old pvdisplay:

--- Physical volume ---

PV Name               /dev/sda6
VG Name               imap_vg
PV Size               261.11 GB [547591527 secs] / NOT usable 16.25 MB [LVM: 189 KB]
PV#                   1
PV Status             NOT available
Allocatable           yes (but full)
Cur LV                1
PE Size (KByte)       16384
Total PE              16710
Free PE               0
Allocated PE          16710
PV UUID               O7WnwM-5ASt-AsiV-8BlZ-h6KP-NjFl-cdY33C

Patched pvdisplay:

--- Physical volume ---

PV Name               /dev/sda6
VG Name               imap_vg
PV Size               261.11 GB [547591527 secs] / NOT usable 16.25 MB [LVM: 189 KB]
PV#                   1
PV Status             available
Allocatable           yes (but full)
Cur LV                1
PE Size (KByte)       16384
Total PE              16710
Free PE               0
Allocated PE          16710
PV UUID               O7WnwM-5ASt-AsiV-8BlZ-h6KP-NjFl-cdY33C

Quoted 88 lines · Heinz J . Mauelshagen

Quoted 13 lines · Rupert Hair
Quoted 2 lines
do all 3 PVs show up active (the [] following 'PV:' should contain 'AA' unless you set a PV to inallocatable) in /proc/lvm/global ?
LVM driver LVM version 1.0.5(15/07/2002)
Total: 1 VG 3 PVs 1 LV (1 LV open 1 times)
Global: 179278 bytes malloced IOP version: 10 5 days 1:33:43 active
VG: datavg [3 PV, 1 LV/1 open] PE Size: 32768 KB
Usage [KB/PE]: 357171200 /10900 total 357171200 /10900 used 0 /0 free
PVs: [AA] hdg1 119996416 /3662 119996416 /3662
0 /0
Quoted 7 lines
      [AA] hde1                  119996416 /3662    119996416 /3662
0 /0
      [AA] hdf1                  117178368 /3576    117178368 /3576
0 /0
LV: [AWDL ] datalv 357171200 /10900 1x open
Looks like Patrick and I have the same problem. Thanks for your help Heinz.
Hmmm, pretty od really. pvdisplay calls the pv_status() lib function which retrieves the information from the kernel. That's why i asked for the proc output which shows that the PVs are _active_ and allocatable.
Wait, need to preserve the core flags to make this work...
Could you test if the following patch to pvdisplay.c helps ?
diff -u -B -u -r1.25 pvdisplay.c
pvdisplay.c 11 Feb 2003 15:31:51 -0000 renamed from pvdisplay.c 6 Feb 2003 15:05:30 -0000 1.25 +3 -1
@@ -42,6 +42,7 @@
  *    06/02/2002 - cope with changed pv_get_size() return cast
  *    07/02/2002 - fixes for > 1TB support
  *    17/02/2002 - avoid pv_get_size() altogether
+ *    11/02/2002 - fixed bug preventing correct 'active' status displayed [HM]
  *
  */
 
@@ -231,7 +232,8 @@
          return LVM_EPVDISPLAY_PV_CHECK_CONSISTENCY;
       }
 
-      /* restore the from disk pointer */
+      /* restore the from disk pointer _but_ preserve kernel status */
+      pv_from_disk->pv_status = pv->pv_status;
       pv = pv_from_disk;
 
       if ( opt_c == 0) {
Quoted 7 lines
Rupert
Quoted 5 lines
_______________________________________________
linux-lvm mailing list
[email protected]
http://lists.sistina.com/mailman/listinfo/linux-lvm
read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
Regards,
Heinz -- The LVM Guy --
*** Software bugs are stupid.
Nevertheless it needs not so stupid people to solve them ***
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Heinz Mauelshagen Sistina Software Inc.
Senior Consultant/Developer Am Sonnenhang 11
56242 Marienrachdorf Germany
[email protected] +49 2626 141200
                                                       FAX 924446
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
_______________________________________________
linux-lvm mailing list
[email protected]
http://lists.sistina.com/mailman/listinfo/linux-lvm
read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/

Rupert Hair

author

2003-02-11 16:47 UTC #6 lore raw

Quoted 1 line

Works for me.

Same here.

Thanks again Heinz.

Rupert

Heinz J . Mauelshagen

2003-02-13 06:09 UTC #7 lore raw

Quoted 2 lines · Rupert Hair

Quoted 1 line
Works for me.
Same here.

Quoted 12 lines

Thanks again Heinz.
Rupert
_______________________________________________
linux-lvm mailing list
[email protected]
http://lists.sistina.com/mailman/listinfo/linux-lvm
read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/

Signature · 15 lines

Regards,
Heinz    -- The LVM Guy --

*** Software bugs are stupid.
    Nevertheless it needs not so stupid people to solve them ***

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Heinz Mauelshagen                                 Sistina Software Inc.
Senior Consultant/Developer                       Am Sonnenhang 11
                                                  56242 Marienrachdorf
                                                  Germany
[email protected]                           +49 2626 141200
                                                       FAX 924446
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-