cancel
Showing results for 
Search instead for 
Did you mean: 

VxVM 4.1 SAN migrate at host level Mirrored volumes are associated with DRL

ramesh_sundar1
Level 2
Partner Accredited Certified

Hi,

I have done SAN migrate earlier without DRL & Subvolume setup. It's mirroed between 2 arrays at host level using vxvm. But now I'm in a position where I need to migrate 100+ volumes between different new arrays. Some volumes have got DRL(log only plexes) & few subvolumes. Unfortunately no documentation of how these DRL(log only plexes) & few subvolume can be migrated. so I'm looking for some help.

The cluster is:

- Solaris 10
- Solaris Cluster 3.1 two node campus cluster
- VxVM 4.1 with CVM
- 2 * NetApp storages mirroed at host level using VxVM

Migration Plan questions is:

 

- 2 * new fujitsu storages have installed & created luns & mapped to hosts on both as per existing Netapp storages
- All new luns detected correctly under OS & VxVM
- I have 4 way mirroing at host level using VxVM for some of volumes.

how can I proceed for DRL(log only plexes) & few subvolume volumes? 


Sample output:-

v  lunx3     -            ENABLED  ACTIVE   2097152  SELECT    -        fsgen
pl lunx3-04  lunx3     ENABLED  ACTIVE   2097152  CONCAT    -        RW
sv lunx3-S01 lunx3-04  lunx3-L01 1    2097152  0         3/3      ENA                 => subvolume

v  lunx3-L01 -            ENABLED  ACTIVE   2097152  SELECT    -        fsgen
pl lunx3-P01 lunx3-L01 ENABLED  ACTIVE   LOGONLY  CONCAT    -        RW                => LOGONLY plex
sd netapp2_datavol-64 lunx3-P01 netapp2_datavol 157523968 528 LOG  FAS60400_0 ENA
pl lunx3-P02 lunx3-L01 ENABLED  ACTIVE   2097152  CONCAT    -        RW
sd netapp2_datavol-65 lunx3-P02 netapp2_datavol 157524496 2097152 0 FAS60400_0 ENA
pl lunx3-P03 lunx3-L01 ENABLED  ACTIVE   2097152  CONCAT    -        RW
sd netapp1_datavol-63 lunx3-P03 netapp1_datavol 157523968 2097152 0 FAS60401_5 ENA

v  lunx4     -            ENABLED  ACTIVE   2097152  SELECT    -        fsgen
pl lunx4-04  lunx4     ENABLED  ACTIVE   2097152  CONCAT    -        RW
sv lunx4-S01 lunx4-04  lunx4-L01 1    2097152  0         3/3      ENA                => subvolume

v  lunx4-L01 -            ENABLED  ACTIVE   2097152  SELECT    -        fsgen
pl lunx4-P01 lunx4-L01 ENABLED  ACTIVE   LOGONLY  CONCAT    -        RW                => LOGONLY plex
sd netapp1_datavol-64 lunx4-P01 netapp1_datavol 161718272 528 LOG  FAS60401_5 ENA
pl lunx4-P02 lunx4-L01 ENABLED  ACTIVE   2097152  CONCAT    -        RW
sd netapp1_datavol-65 lunx4-P02 netapp1_datavol 159621120 2097152 0 FAS60401_5 ENA
pl lunx4-P03 lunx4-L01 ENABLED  ACTIVE   2097152  CONCAT    -        RW
sd netapp2_datavol-66 lunx4-P03 netapp2_datavol 159622176 2097152 0 FAS60400_0 ENA

 

Thanks in Advance

Ramesh Sundaram

 

 

1 ACCEPTED SOLUTION

Accepted Solutions

Gaurav_S
Moderator
Moderator
   VIP    Certified

Hi Ramesh,

You can simply use the vxassist mirror command as directed in the link I gave & provide new disk name at the end so that mirror plex is taken from new storage ... 

vxassist -g <diskgroup>  mirror <volume> <list_of_new_disks>

That will add 1 mirror copy, once you have confirmed mirror copy, you can disassociate the old plex from old storage using vxassist -g <diskgroup> remove mirror option

Once the volume is surviving on disk from new storage, you can again mirror the volume to add one more copy from new storage..

About DRL, Dirty region logging (DRL), if enabled, speeds recovery of mirrored volumes after a system crash. DRL keeps track of the regions that have changed due to I/O writes to a mirrored volume. DRL uses this information to recover only those portions of the volume that need to be recovered. Once volume is mirrored & available from new storage, you can add DRL log using below command

# vxassist -g <diskgorup> addlog <volume> logtype=drl

 

G

 

View solution in original post

4 REPLIES 4

Gaurav_S
Moderator
Moderator
   VIP    Certified

Hi Ramesh,

First suggestion, vxvm is at very old version, would suggest strongly to upgrade to latest versions ..

For the target host, do you intend to keep the same structure of volume ? this is not mandatory. that means if your current source volume is a layered volume (stripe), you can change the construct of volume on target Lun (mirror or stripe). You should also consider the use of the volume, if the volume is a write intensive volume, you may want to keep volume striped.

Another choice would be to convert the volume at source itself to a simple structure & then mirror it with disk from new fujitsu storage.

DRL log can be removed & reattached to a volume. this would not impact to the data of the volume.

Refer this post, it may help

https://www-secure.symantec.com/connect/forums/veritas-layed-volume-mirroring

 

G

 

ramesh_sundar1
Level 2
Partner Accredited Certified

Hi Gaurav,

Thanks for your update. We don't want subvolume.. we want normal volume with mirroed plexes as following example

v  vol139       -            ENABLED  ACTIVE   2097152  SELECT    -        fsgen
pl vol139-01    vol139       ENABLED  ACTIVE   2097152  CONCAT    -        RW
sd fuji3_datamcs-60 vol139-01 fuji3_datamcs 153329664 2097152 0   Disk_35 ENA
pl vol139-02    vol139       ENABLED  ACTIVE   2097152  CONCAT    -        RW
sd fuj4_datamcs-60 vol139-02 fuji4_datamcs 153329664 2097152 0   Disk_34 ENA

1) How it can be done? will there be any data corruption risk?

2) Do we realy need LOGONLY plex? If we remove LOGONLY plex permanatly what would happen?

Regards, Ramesh S

 

 

 

 

Gaurav_S
Moderator
Moderator
   VIP    Certified

Hi Ramesh,

You can simply use the vxassist mirror command as directed in the link I gave & provide new disk name at the end so that mirror plex is taken from new storage ... 

vxassist -g <diskgroup>  mirror <volume> <list_of_new_disks>

That will add 1 mirror copy, once you have confirmed mirror copy, you can disassociate the old plex from old storage using vxassist -g <diskgroup> remove mirror option

Once the volume is surviving on disk from new storage, you can again mirror the volume to add one more copy from new storage..

About DRL, Dirty region logging (DRL), if enabled, speeds recovery of mirrored volumes after a system crash. DRL keeps track of the regions that have changed due to I/O writes to a mirrored volume. DRL uses this information to recover only those portions of the volume that need to be recovered. Once volume is mirrored & available from new storage, you can add DRL log using below command

# vxassist -g <diskgorup> addlog <volume> logtype=drl

 

G

 

ramesh_sundar1
Level 2
Partner Accredited Certified

Hi Gaurav,

Thanks a lot.

Regards, Ramesh S