The error results because there is at least one active volume using the disk being removed. e.g.

dg testdg       default      default  28000    1641672206.15.localhost.localdomain

dm vrts_appliances0_4 vrts_appliances0_4 auto 65536 104791760 -
dm vrts_appliances0_5 vrts_appliances0_5 auto 65536 104791760 -

v  testvol2     -            ENABLED  ACTIVE   204800   SELECT    -        fsgen
pl testvol2-01  testvol2     ENABLED  ACTIVE   204800   CONCAT    -        RW
sd vrts_appliances0_5-01 testvol2-01 vrts_appliances0_5 0 204800 0 vrts_appliances0_5 ENA
pl testvol2-02  testvol2     ENABLED  ACTIVE   204800   CONCAT    -        RW
sd vrts_appliances0_4-02 testvol2-02 vrts_appliances0_4 204800 204800 0 vrts_appliances0_4 ENA

vrts_appliances0_4 is occupied by a data plex by the testvol2 volume

With the given scenario, a decision was taken to use the -k option. This put the LUN in a removed state:

e.g.

# vxdg -g testdg -k rmdisk vrts_appliances0_4
dm vrts_appliances0_4 -      -        -        -        REMOVED
dm vrts_appliances0_5 vrts_appliances0_5 auto 65536 104791760 -

v  testvol2     -            ENABLED  ACTIVE   204800   SELECT    -        fsgen
pl testvol2-01  testvol2     ENABLED  ACTIVE   204800   CONCAT    -        RW
sd vrts_appliances0_5-01 testvol2-01 vrts_appliances0_5 0 204800 0 vrts_appliances0_5 ENA
pl testvol2-02  testvol2     DISABLED REMOVED  204800   CONCAT    -        RW
sd vrts_appliances0_4-02 testvol2-02 vrts_appliances0_4 204800 204800 0 -  RMOV
         
The vrts_appliances0_4 LUN was now in REMOVED state. The volume was still ENABLED:ACTIVE as one plex was still enabled. It is not recommended that the -k option be used unless the data happens to have no importance.

Return the removed LUN back to its former state (and remove any unwanted data plexes first):

# vxdg -g testdg -k adddisk vrts_appliances0_4

dm vrts_appliances0_4 vrts_appliances0_4 auto 65536 104791760 -
dm vrts_appliances0_5 vrts_appliances0_5 auto 65536 104791760 -

v  testvol2     -            ENABLED  ACTIVE   204800   SELECT    -        fsgen
pl testvol2-01  testvol2     ENABLED  ACTIVE   204800   CONCAT    -        RW
sd vrts_appliances0_5-01 testvol2-01 vrts_appliances0_5 0 204800 0 vrts_appliances0_5 ENA
pl testvol2-02  testvol2     DISABLED RECOVER  204800   CONCAT    -        RW
sd vrts_appliances0_4-02 testvol2-02 vrts_appliances0_4 204800 204800 0 vrts_appliances0_4 ENA

The disk had been added back and plexvol1-01 was now in DISABLED RECOVER state. Running vxrecover resyncs the plex:

# vxrecover -g testdg testvol2

dm vrts_appliances0_4 vrts_appliances0_4 auto 65536 104791760 -
dm vrts_appliances0_5 vrts_appliances0_5 auto 65536 104791760 -

v  testvol2     -            ENABLED  ACTIVE   204800   SELECT    -        fsgen
pl testvol2-01  testvol2     ENABLED  ACTIVE   204800   CONCAT    -        RW
sd vrts_appliances0_5-01 testvol2-01 vrts_appliances0_5 0 204800 0 vrts_appliances0_5 ENA
pl testvol2-02  testvol2     ENABLED  ACTIVE   204800   CONCAT    -        RW
sd vrts_appliances0_4-02 testvol2-02 vrts_appliances0_4 204800 204800 0 vrts_appliances0_4 ENA

The LUN can now be correctly removed by first disassociating and removing any plexes on the LUN.

# vxplex -g testdg -o rm dis testvol2-02
dm vrts_appliances0_4 vrts_appliances0_4 auto 65536 104791760 -
dm vrts_appliances0_5 vrts_appliances0_5 auto 65536 104791760 -

v  testvol2     -            ENABLED  ACTIVE   204800   SELECT    -        fsgen
pl testvol2-01  testvol2     ENABLED  ACTIVE   204800   CONCAT    -        RW
sd vrts_appliances0_5-01 testvol2-01 vrts_appliances0_5 0 204800 0 vrts_appliances0_5 ENA

# vxdg -g testdg rmdisk vrts_appliances0_4
dm vrts_appliances0_5 vrts_appliances0_5 auto 65536 104791760 -

v  testvol2     -            ENABLED  ACTIVE   204800   SELECT    -        fsgen
pl testvol2-01  testvol2     ENABLED  ACTIVE   204800   CONCAT    -        RW
sd vrts_appliances0_5-01 testvol2-01 vrts_appliances0_5 0 204800 0 vrts_appliances0_5 ENA