Tuesday, July 3, 2018

Unable to open device "/dev/sdab1": Device or resource busy during oracleasm disk creation.


Resource or device busy issue when creating the ASM disks at OS level usign RAW disks provided from NETAPP Storage.


Issue:- Oracle ASM disks labeling failed for one of the Multi path RAW device which was provided from NETAPP storage as below

All the provided LUNs got mounted on DB Servers using below commands.

echo "- - -" > /sys/class/scsi_host/host1/scan
echo "- - -" > /sys/class/scsi_host/host2/scan

Once the disks got appeared at OS level in both nodes (2-Node RAC DB Server) the actual issue encountered.

[root@ebsdbser01 mapper]# fdisk -l /dev/sds
Disk /dev/sds: 214.7 GB, 214748364800 bytes
255 heads, 63 sectors/track, 26108 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 65536 bytes
Disk identifier: 0x40402c59

   Device Boot      Start         End      Blocks   Id  System
/dev/sds1               1       26108   209712478+  83  Linux
Partition 1 does not start on physical sector boundary.

[root@ebsdbser01 mapper]# fdisk -l /dev/sdab

Disk /dev/sdab: 214.7 GB, 214748364800 bytes
255 heads, 63 sectors/track, 26108 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 65536 bytes
Disk identifier: 0x40402c59

    Device Boot      Start         End      Blocks   Id  System
/dev/sdab1               1       26108   209712478+  83  Linux
Partition 1 does not start on physical sector boundary.

[root@ebsdbser01 mapper]# oracleasm createdisk FLASH01 /dev/sdab1

Unable to open device "/dev/sdab1": Device or resource busy



Work Around:-

As a work around, the Mapper device was taken for logical partitioning (fdisk) instead of its related one of the raw device path and then partprobe was used after partitioned. Finally all the dependent mutipaths got logically partitioned. And the ASM DISKS labelling successfully completed on both 2 node DB servers.

[root@ebsdbser01 mapper]# multipath -ll mpathe
mpathe (3600a098038303778462b49533170412f) dm-14 NETAPP,LUN C-Mode
size=200G features='4 queue_if_no_path pg_init_retries 50 retain_attached_hw_handle' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=50 status=active
| |- 1:0:1:200 sds  65:32  active ready running
| `- 2:0:0:200 sdab 65:176 active ready running
`-+- policy='round-robin 0' prio=10 status=enabled
  |- 1:0:0:200 sdj  8:144  active ready running
  `- 2:0:1:200 sdak 66:64  active ready running
 
[root@ebsdbser01 mapper]# partprobe /dev/mapper/mpathe

[root@ebsdbser01 mapper]# ls -ltr /dev/mapper/mpathe*
lrwxrwxrwx. 1 root root 8 Jul  3 15:46 /dev/mapper/mpathep1 -> ../dm-24
lrwxrwxrwx. 1 root root 8 Jul  3 15:46 /dev/mapper/mpathe -> ../dm-14

[root@ebsdbser01 mapper]# oracleasm createdisk FLASH01 /dev/mapper/mpathep1
Writing disk header: done
Instantiating disk: done

[root@ebsdbser01 mapper]# oracleasm querydisk -d FLASH01
Disk "FLASH01" is a valid ASM disk on device [251,24]

[root@ebsdbser01 mapper]# oracleasm querydisk -p FLASH01
Disk "FLASH01" is a valid ASM disk
/dev/sdj1: LABEL="FLASH01" TYPE="oracleasm"
/dev/sds1: LABEL="FLASH01" TYPE="oracleasm"
/dev/sdab1: LABEL="FLASH01" TYPE="oracleasm"
/dev/sdak1: LABEL="FLASH01" TYPE="oracleasm"
/dev/mapper/mpathep1: LABEL="FLASH01" TYPE="oracleasm"

No comments:

Post a Comment

Oracle RAC node unavailable with error: Server unexpectedly closed network connection6]clsc_connect: (0x251c670) no listener at (ADDRESS=(PROTOCOL=ipc)(KEY=OCSSD_LL_node2_))

 Early midnight I received a call from the monitoring team that one of the critical production database node is not available. As I am aware...