vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
NAME [Toc] [Back]
vradmin - administer VERITAS Volume Replicator (VVR) in a distributed
environment
SYNOPSIS [Toc] [Back]
vradmin [-g diskgroup] [-nodcm] [-sdg diskgroup] addsec rvg prihost
sechost [attribute...]
vradmin [-g diskgroup] [-nodcm] [-s] addvol rvg volume
vradmin [-g diskgroup] changeip rvg [sechost] [attribute...]
vradmin [-g diskgroup] [-nodcm] createpri rvg volumelist srl
vradmin [-g diskgroup] [-f] delpri rvg
vradmin [-g diskgroup] delsec rvg [sechost]
vradmin [-g diskgroup] [-f] delvol rvg volume
vradmin [-g diskgroup] [-wait] fbsync rvg
vradmin [-g diskgroup] [-all] ibc rvg taskname [sechost] ...
vradmin [-g diskgroup] makesec rvg prihost
vradmin [-g diskgroup] [-s] migrate rvg [sechost]
vradmin [-g diskgroup] pauserep rvg [sechost]
vradmin [-g diskgroup] [-l] printrvg [rvg...]
vradmin [-g diskgroup] printvol [rvg...]
vradmin [-g diskgroup] [-l] repstatus rvg
vradmin [-g diskgroup] [-f] resizevol rvg volume length
vradmin [-g diskgroup] resumerep rvg [sechost]
vradmin [-g diskgroup] resync rvg
vradmin [-g diskgroup] set rvg [sechost] attribute...
vradmin [-g diskgroup] { -a | -c checkpoint | -f } startrep rvg
[sechost]
vradmin [-g diskgroup] [-f] [-s] stoprep rvg [sechost]
vradmin [-g diskgroup] -c checkpoint [-full] syncrvg rvg sechost...
- 1 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
vradmin [-g diskgroup] [-s] -verify syncrvg rvg sechost...
vradmin [-g diskgroup] [-full] [-s] syncvol volumelist
host[:diskgroup[:volumelist]]...
vradmin [-g diskgroup] [-s] -verify syncvol volumelist
host[:diskgroup[:volumelist]]...
vradmin [-g diskgroup] [-autofb] [-N] takeover rvg
vradmin help [keyword]
vradmin -version
DESCRIPTION [Toc] [Back]
The vradmin utility performs administrative tasks on a Replicated Data
Set (RDS) in a VERITAS Volume Replicator (VVR) environment. Tasks
available in vradmin include configuring, administering and displaying
Replicated Data Sets, synchronizing remote volumes, and migrating or
taking over the Primary role in the RDS.
Note: vradmin is a VVR-specific utility and requires a valid VVR
license.
A Replicated Data Set (RDS) is comprised of a Replicated Volume Group
(RVG) on the Primary host and its counterparts on the Secondary hosts.
Note: An RDS is a concept used in vradmin; it is not a VVR object.
To perform an administrative task, vradmin accesses and changes the
VVR configuration on the Primary and the Secondary hosts in an RDS.
The vradmin command can be executed from a Primary or any Secondary
host in an RDS.
The vradmin keywords specify the operation to perform. Each operation
can be performed on only one RDS at a time. The rvg argument is the
name of a Replicated Volume Group (RVG) that specifies the RDS on the
local host. Because the name of the Primary and Secondary RVG can be
different, the same RDS can be represented by different names on
different hosts.
The rvg or volume arguments determine a default local disk group,
according to the standard disk group selection rules described in
vxintro(1M). You can specify a local disk group using the -g
diskgroup option.
KEYWORDS [Toc] [Back]
addsec Creates a Secondary RVG of the same name as the Primary RVG
and adds it to the Replicated Data Set (RDS) rvg. The
addsec command also associates existing data volumes and
- 2 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
Storage Replicator Log (SRL) on the Secondary with the
Secondary RVG. It creates and associates Primary and
Secondary RLINKs (named rlk_sechost_rvg and rlk_prihost_rvg)
with the Primary and Secondary RVGs.
The addsec command requires that the data volumes and SRL
must exist on the Secondary with the same names and lengths
as the data volumes and SRL on the Primary.
The arguments prihost and sechost specify the network
connection to use for replication. They are resolvable
names or IP addresses of the Primary and the Secondary host.
The addsec command can be executed from any host in the RDS.
If an RDS contains only a Primary host, then addsec must be
executed from the Primary host.
Caution: Do not execute the addsec command from the
Secondary host that is being added to the RDS.
The Secondary can be created only if the /etc/vx/vras/.rdg
file on the Secondary host contains an entry for the Primary
disk group ID. Before creating the Secondary, this entry is
validated as part of the addsec command. A plus character
(+) entry in the /etc/vx/vras/.rdg file on the Secondary
host allows addsec command for any disk group ID from any
remote host to proceed.
The addsec command creates a Secondary RVG in the disk group
with the same name as the Primary disk group. If required,
you can use the option -sdg diskgroup to specify a different
disk group on the Secondary.
If any of the data volumes or the SRL on the Secondary has
Dirty Region Logs (DRL), the DRL is removed before the
volume is associated with the Secondary RVG.
By default, the addsec command adds Data Change Maps (DCM)
to the Primary and Secondary data volumes. Use the -nodcm
option to specify that DCMs are not to be added to the data
volumes. When the -nodcm option is used, the addsec command
sets the srlprot attribute of the RLINKs to the value off
instead of the default value of autodcm.
By default, the addsec command creates RLINKs named
rlk_remotehost_rvgname. You can specify Primary and
Secondary RLINK names using the prlink and srlink
attributes. The expression to assign a value to an attribute
takes the form attribute_name=value. For example, the
following command creates Primary RLINK to_nyk and Secondary
RLINK to_sfo, and adds Secondary nyk to RDS hr_rvg:
- 3 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
vradmin addsec hr_rvg sfo nyk prlink=to_nyk srlink=to_sfo
addvol Adds an existing volume to the Replicated Data Set (RDS)
rvg. This command associates an existing volume with the
corresponding RVGs on all hosts in the RDS. The volume must
have the same length on the Primary and Secondary hosts in
the RDS.
If any of the Primary or Secondary data volumes that are to
be added to the RDS have Dirty Region Logs (DRL), the addvol
command removes the DRLs from these data volumes.
By default, the addvol command adds Data Change Maps (DCM)
to the Primary and Secondary data volumes. Use the -nodcm
option if you do not want to add DCM to the volumes.
If replication is in progress, make sure that the Secondary
volume contains the same data as the Primary volume before
adding the volume to the RDS. To transfer the data from the
Primary volume to the Secondary volume over the network, use
the vradmin syncvol command.
At the start of the addvol command, you are asked to confirm
that the Primary and Secondary volumes contain the same
data. Specifying the -s option skips this confirmation
step.
changeip Changes the network used for the replication between the
Primary and Secondary sechost in the Replicated Data Set
(RDS) rvg. The argument sechost is the name of the
Secondary host as displayed by the vradmin printrvg command.
The argument sechost is optional if the RDS contains only
one Secondary host.
The vradmin changeip command changes the hostname or IP
address set in the Primary and Secondary RLINKs to the new
values specified in the newpri and newsec attributes. These
attributes are of the form attribute_name=value.
newpri
Specifies a new hostname or IP address for the Primary
host that is to be used to establish a network
connection for the replication to the Secondary. This
is the new value for the local_host attribute of the
Primary RLINK and the remote_host attribute of the
corresponding Secondary RLINK.
- 4 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
newsec
Specifies a new hostname or IP address for the
Secondary host that is to be used to establish a
network connection for the replication. This is the
new value for the remote_host attribute of the Primary
RLINK and the local_host attribute of the Secondary
RLINK.
Refer to vxedit(1M) for more information on the local_host
and remote_host attributes of an RLINK.
Note: The new hostname or IP address of the Primary and
Secondary must be resolvable and at least one of the
original or new network connections must be available.
createpri Creates a Primary RVG rvg for the new Replicated Data Set
(RDS) rvg. The volumelist argument is a comma-separated
list of the existing data volumes. The srl argument is the
existing volume that is to be used as the Storage Replicator
Log (SRL) volume.
Before using the createpri command, use the vxassist or the
vxmake command to create the data volumes and SRL with the
required layout. Execute the createpri command on the host
that is to be configured as the Primary host of the new RDS.
The createpri command starts the Primary RVG and then
associates the data volumes and the SRL volume with the
Primary RVG.
The createpri command removes the Dirty Region Logs (DRL)
from the data volumes and the SRL volume if they have DRLs.
By default, the createpri command adds Data Change Maps
(DCM) to the data volumes. Use the -nodcm option if you do
not want to add DCM to the data volumes.
Note: You can run the createpri command when the data
volumes are in use (active).
delpri Deletes the Primary RVG (rvg), thereby deleting the
corresponding Replicated Data Set (RDS) rvg. This command
dissociates the data volumes and the SRL from the Primary
RVG; it does not delete the data volumes and the SRL from
the VERITAS Volume Manager configuration.
The delpri command fails if the Primary RVG to be deleted
still has one or more Secondaries. Use the vradmin delsec
command to delete all of its Secondaries before using the
delpri command to delete the Primary RVG.
- 5 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
Use the -f option to delete a Primary RVG when the Primary
data volumes are in use.
This command can be run only from the Primary host.
delsec Deletes the Secondary sechost from the Replicated Data Set
(RDS) rvg. The argument sechost is the name of the
Secondary host as displayed in the output of the vradmin
printrvg command. The argument sechost is optional if the
RDS contains only one Secondary host. The delsec command
dissociates the data volumes and the SRL from the Secondary
RVG and deletes the Secondary RVG, the Secondary RLINK, and
the Primary RLINK to the Secondary.
Note: The delsec command does not delete the data volumes
and SRL.
The delsec command fails if replication to the Secondary
sechost is active. Use the vradmin stoprep command to stop
replication to the Secondary.
delvol Deletes the data volume volume from the Replicated Data Set
(RDS) rvg. The delvol command dissociates the volume from
the Primary and the Secondary RVGs in the RDS. The volumes
are not physically deleted from the VxVM configuration.
You can remove a data volume from its RDS either when the
replication is in progress (that is, RLINKs are in the
CONNECT state), or when the replication is stopped (that is,
RLINKs are detached).
Use the -f option to delete a volume from the RDS when the
Primary RVG is not stopped.
fbsync Converts a failed Primary to a Secondary in the Replicated
Data Set (RDS) rvg. It also uses failback logging to
synchronize data volumes on the failed Primary with the data
volumes on the new Primary.
Note: This command can be run only if failback logging was
enabled during the takeover. Refer to the takeover command
for more details on enabling the failback logging.
In failback logging mode, VVR uses Data Change Maps (DCM) to
track the changes happening on the new Primary while the
original Primary is not available. The fbsync command
synchronizes the original Primary with the new changes on
the new Primary by replaying the DCMs.
This method of failed Primary synchronization is recommended
over the difference-based synchronization that is used by
- 6 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
the syncrvg command. Failback logging neither requires all
data blocks to be read, nor computation and comparison of
checksums for the data blocks to be performed.
The fbsync command returns immediately after starting the
synchronization. Use the -wait option if you want the
fbsync command to wait until the synchronization is
complete.
ibc Performs off-host processing task on the Secondary sechost
in the Replicated Data Set (RDS) rvg using In-Band Control
(IBC) Messaging. The argument sechost is the name of the
Secondary host as displayed in the output of the vradmin
printrvg command. The argument sechost is optional if the
RDS contains only one Secondary host. To perform the task on
multiple Secondaries, specify a space-separated list of
their names or IP addresses. Use the -all option to perform
the task on all Secondaries in the RDS. The argument
taskname identifies the off-host processing task and the
corresponding user-defined scripts.
The vradmin ibc command uses IBC Messaging and executes a
set of user-defined scripts to perform certain desired offhost
processing task. See the VERITAS Volume Replicator
Administrator's Guide for more information on In-Band
Control Messaging.
The vradmin ibc command fails if the RLINKs are not in
CONNECT state.
The vradmin ibc command executes the following user-defined
scripts:
On the Primary host:
/etc/vx/vvr/ibc_scripts/<taskname>/quiesce
/etc/vx/vvr/ibc_scripts/<taskname>/unquiesce
On the Secondary sechost host:
/etc/vx/vvr/ibc_scripts/<taskname>/prefreeze
/etc/vx/vvr/ibc_scripts/<taskname>/onfreeze
/etc/vx/vvr/ibc_scripts/<taskname>/postfreeze
The onfreeze script is mandatory. The vradmin ibc command
fails if the onfreeze script does not exist on the Secondary
host. The other scripts are optional. If the quiesce script
is provided, the unquiesce script must also be provided and
vice versa.
Below is a brief guideline on which operations should be
implemented in each script:
- 7 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
quiesce:
Quiesce the application running on the Primary data
volumes and put the Primary data volumes in
application-consistent state.
unquiesce:
Unquiesce the application running on the Primary data
volumes, i.e. put the application back to the normal
operating mode.
prefreeze:
Perform operations needed on the Secondary RVG in
preparation for the execution of the onfreeze script.
For example, if the action in the onfreeze script is to
take a snapshot of the Secondary data volumes, the
prefreeze script can be used to add snapshot plexes to
the Secondary data volumes.
onfreeze:
Perform desired operation that requires the replication
to be frozen on the Secondary RVG such as taking a
snapshot of the Secondary data volumes.
postfreeze:
Perform any neccessary operations after the onfreeze
script has completed and the replication on the
Secondary RVG has resumed.
See the VERITAS Volume Replicator Administrator's Guide for
detailed information on how these scripts work in the
vradmin ibc command.
makesec Converts a failed Primary to a Secondary of the new Primary
in the Replicated Data Set (RDS) rvg. The argument prihost
is the name of the new Primary host that was previously a
Secondary of the failed Primary. The vradmin -l printrvg rvg
command displays Primary-Primary configuration error if
there are multiple Primary nodes in the Replicated Data Set
(RDS) rvg. The prihost argument must be the host that is
displayed with a Primary-Primary configuration error.
Note: This command can be run only from failed Primary host
when one of its Secondaries has taken over the Primary role.
Use the makesec command to convert a failed Primary RVG to a
Secondary in a failback procedure only if failback logging
was not enabled during the takeover. To perform failback,
first use the vradmin makesec command to convert the failed
Primary to a Secondary. Then use the vradmin syncrvg command
to synchronize the data volumes on the new Secondary (failed
Primary) with the data volumes on the new Primary. Finally,
- 8 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
at a scheduled time, use the vradmin migrate command to
migrate the Primary role back to the failed Primary.
migrate Migrates the Primary role from the Primary host to a
Secondary host in the Replicated Data Set (RDS) rvg. The
Primary role can only be migrated when all the attached
Primary RLINKs are up-to-date, consistent, and not in an
error state. The data volumes in the RDS must be inactive,
that is, applications involved in replication must be
stopped before running the migrate command. After the
migration completes, the former Secondary adopts the Primary
role, and the former Primary adopts the Secondary role.
By default, the migrate command starts replication from the
new Primary to the original Primary, which is now a
Secondary. The command issues a warning if the replication
does not start (i.e. RLINKs do not change to CONNECT state).
The migrate command also starts the RVG on the new Primary.
At the start of the migrate command, you are asked to
confirm that all the applications using the Primary volumes
have been stopped. You can use the -s option to skip this
confirmation step.
pauserep Pauses the replication to the Secondary sechost in the
Replicated Data Set (RDS) rvg. The argument sechost is the
name of the Secondary host as displayed in the output of
vradmin printrvg command. The argument sechost is optional
if the RDS contains only one Secondary host.
The pauserep command initiates a Primary pause that is
independent of the host from which the command is executed.
See the vxrlink(1M) manual page for more information on
Primary pause and Secondary pause.
printrvg Displays information about all RVGs in the Replicated Data
Set (RDS) rvg. The host name, the RVG name, and the disk
group are displayed for each RVG in the RDS starting with
the Primary RVG. If the argument rvg is not specified, all
Replicated Data Sets on the local host are displayed. Use
the -g diskgroup option to display Replicated Data Sets in a
particular disk group.
The -l option displays information in a long format. This
format displays additional information such as number of
data volumes, SRL name, and RLINK-related information for
each RVG in the RDS. This option also displays configuration
errors in the RDS, if there are any.
printvol Displays information about data volumes in all RVGs in the
Replicated Data Set (RDS) rvg. The host, disk group, RVG,
- 9 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
and volume names are displayed for each RVG in the RDS
starting with the Primary RVG. If the argument rvg is not
specified, the above information is displayed for all
Replicated Data Sets on the local host.
You can use the -g diskgroup option to display Replicated
Data Sets in a particular disk group.
Information is displayed in a table format. The first
column is the type field for each row. The second column
contains information about the Primary RVG, if there is one.
Information about the Secondary hosts is displayed in next
columns. If there are more than two Secondary hosts in the
RDS, information tables are repeated with the Primary in the
second column.
The name, size (in blocks), plex count, and volume flags are
displayed for each volume in the RDS. The printvol command
is useful to view volume mappings between the Primary and
the Secondary RVGs. Secondary volumes are displayed on the
same row as the corresponding Primary volumes in the output.
If a Primary or Secondary volume is not mapped, a dash (-)
is displayed.
repstatus Displays the current replication status of the Replicated
Data Set (RDS) rvg. repstatus displays information about
each RVG in the RDS starting with the Primary RVG.
Output of the repstatus command contains the following
information for the Primary RVG: host name, RVG name, disk
group, RVG state, number of data volumes being replicated,
SRL name, SRL size, and the number of Secondary RVGs
configured in the RDS.
For each Secondary RVG in the RDS, the following information
displays: host name, RVG name, disk group, data status,
replication status, current replication mode, and type of
log currently in use. If the -l option is used, additional
information such as RLINK names, replication mode configured
in the RLINKs, latency protection setting, and SRL
protection setting is also displayed for all the Secondary
RVGs.
Following is a description of the important fields in the
output of repstatus command:
RVG state [Toc] [Back]
Indicates the state of the Primary RVG, for example,
Primary RVG is currently enabled or disable for I/O.
- 10 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
Data status [Toc] [Back]
Indicates whether the data on this Secondary is
consistent or inconsistent with respect to the data on
the Primary. If the Secondary is consistent, repstatus
also displays whether the Secondary data is up-to-date,
behind, or in stale state. A value of N/A indicates if
the Secondary data needs recovery or its status cannot
be determined (this may be due to a configuration
error).
Current replication mode [Toc] [Back]
Displays the current replication mode that is used to
replicate data to the Secondary. This value can be
different than the replication mode configured in the
Primary RLINK if the replication mode configured is
synchronous override.
Replication status [Toc] [Back]
Indicates whether Primary data is being replicated to
the Secondary. repstatus displays additional
information in case data is not replicated.
Logging to [Toc] [Back]
Displays whether updates to the Secondary are tracked
using SRL or DCM. If the Secondary is not up-to-date,
repstatus shows how far behind the Secondary is with
respect to the Primary and the percentage of SRL used
by this Secondary.
The repstatus command also displays configuration errors in
the RDS, if there are any.
Note: if the vradmin repstatus command is run from a
Secondary which has some configuration error or cannot reach
to the Primary, the output of the repstatus command displays
the status known to the Secondary before the above condition
occurs and thus might be out-of-date.
resizevol Resizes the data volume, volume, in a Replicated Data Set
(RDS) rvg to the size specified by length. The resizevol
command resizes the Primary data volume volume and the
corresponding Secondary data volumes.
The argument length is the desired size for the data volume.
The length can be specified using the standard length
convention in VxVM. If the sizes of the Primary and
Secondary data volumes are currently the same, the argument
length can be prefixed by either plus (+) or minus (-) to
indicate whether to increase or decrease the data volume
size by the specified amount.
- 11 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
The resizevol command also resizes the file system on the
Primary data volume as described in vxresize(1M).
If replication to any Secondary RVG is in progress, the
resizevol command performs a secondary pause, resizes the
volume, and then resumes the replication.
Note: Before increasing a volume size, make sure there is
enough free space on the Primary and all the Secondaries to
accommodate the growth, and also make sure that the
additional portion of the Secondary data volumes is
synchronized with the additional portion of the Primary data
volume.
The -f option is required if any of the data volumes to be
resized are being decreased in size.
If the resizevol command fails on any of the hosts in the
RDS during its execution, the original volume sizes are not
restored. This results in volume size mismatch between the
Primary and its Secondaries. To correct this mismatch, reissue
the resizevol command with an absolute length.
resumerep Resumes replication to the Secondary sechost in the
Replicated Data Set (RDS) rvg. The argument sechost is the
name of the Secondary host as displayed in the output of the
vradmin printrvg command. The argument sechost is optional
if the RDS contains only one Secondary host.
The resumerep command resumes the Primary RLINK and the
corresponding Secondary RLINK to resume replication. The
command issues a warning if the replication does not resume
(i.e. RLINKs do not change to CONNECT state).
resync Starts replay an active DCM in the Replicated Data Set
(RDS) rvg.
The resync command starts replay of a DCM that is active due
to SRL overflow or failback logging. Replay occurs for all
RLINKs on which the dcm_logging flag is set. If any of
these RLINKs are disconnected or paused, resynchronization
is delayed until this condition clears. Detaching a
disconnected or paused RLINK clears the dcm_logging flag,
and allows resynchronization to proceed on any remaining
RLINKs.
Refer to vxrvg(1M) for more information.
set Changes the replication attributes for the replication to
the Secondary sechost in the Replicated Data Set (RDS) rvg.
The argument sechost is the name of the Secondary host as
- 12 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
displayed by the vradmin printrvg command. The argument
sechost is optional if the RDS contains only one Secondary
host.
The vradmin set command changes the replication attributes
on both the Primary and Secondary. These attributes are of
the form attribute_name=value.
The following attributes are available:
latency_high_mark
Sets the maximum number of outstanding requests when
latency protection is enabled.
latency_low_mark
After throttling is enabled, sets the level to which
the number of outstanding requests must drop before
throttling is disabled.
latencyprot
Indicates whether latency protection is enabled for the
RLINK. The attribute may have one of following values:
fail, off, or override.
Refer to vxedit(1M) for more information on these
values.
packet_size
Sets the packet_size value. This is the number of bytes
in a packet that is sent to a secondary RLINK.
srlprot
Indicates whether log protection is enabled for the
RLINK. Log protection prevents an RLINK from
overflowing the SRL, which would cause it to become
stale.
This attribute may have one of following values:
autodcm, dcm, fail, off, or override.
Refer to vxedit(1M) for more information on these
values.
synchronous
Indicates whether replication should operate in
synchronous or asynchronous mode.
This attribute may have one of following values: off or
override.
- 13 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
Refer to vxedit(1M) for more information on these
values.
startrep Starts replication to the Secondary sechost in the
Replicated Data Set (RDS) rvg. The argument sechost is the
name of the Secondary host as displayed in the output of
vradmin printrvg command. The argument sechost is optional
if the RDS contains only one Secondary host.
The startrep command attaches RLINKs on the Secondary and
Primary to start replication.
The startrep command provides the -a, -c checkpoint and -f
options. Only one of these options may be specified with
this command.
The -a option starts autosync of the Secondary data volumes
on command execution.
The -c checkpoint option attaches the Primary RLINK with the
specified checkpoint.
The -f (force) option may be used if the Primary RVG is
empty or if the Secondary data volumes contain the exact
same data as Primary data volumes.
These options work in the same way as the corresponding
options of the vxrlink att command (see vxrlink(1M)).
The startrep command issues a warning if the replication
does not start (i.e. RLINKs do not change to CONNECT state).
stoprep Stops replication to the Secondary sechost in the Replicated
Data Set (RDS) rvg. The argument sechost is the name of the
Secondary host as displayed by the vradmin printrvg command.
The argument sechost is optional if the RDS contains only
one Secondary host.
The stoprep command stops the replication by detaching the
RLINKs on the Primary and the Secondary host.
The stoprep command fails if the Secondary data volumes are
not up-to-date with the Primary data volumes or if any of
the Primary data volumes are open. Use the -f option to stop
replication when the Secondary data volumes are not up-todate
or when any of the Primary data volumes are open.
On invoking the stoprep command, you are asked to confirm
that you want to stop the replication. Stopping replication
can result in the data on the Secondary volumes being out of
date with respect to the Primary volumes. If replication is
- 14 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
stopped, an autosync or difference-based synchronization of
the Secondary volumes with the Primary volumes is required
before replication can be restarted. You can use the -s
option to skip this confirmation step.
syncrvg Synchronizes or verifies data volumes on the host sechost in
the Replicated Data Set (RDS) rvg with its Primary data
volumes. The argument sechost is the name or IP address of
the Secondary host as displayed in the output from vradmin
printrvg. One or more Secondaries can be synchronized or
verified at the same time by specifying a space-separated
list of their names or IP addresses.
All volumes associated with Secondary RVGs are synchronized
or verified with the corresponding Primary volumes. This is
done by transferring the Primary volume data over the
network.
There are two types of synchronization: full synchronization
and difference-based synchronization.
In full synchronization, all data on Primary volumes is
transferred to the corresponding Secondary volumes. Full
synchronization is useful for creating the initial volume
copies.
In difference-based synchronization, the difference between
the Primary and the Secondary volume is first computed, and
then only the blocks that are different are transferred to
the Secondary volumes. The MD5 checksum is used to calculate
the difference between the Primary and the Secondary data
volumes. The MD5 checksum is generated using the MD5
Message-Digest Algorithm from RSA Data Security, Inc.
By default, syncrvg and syncvol use difference-based
synchronization. Specify the -full option to perform full
synchronization of the Secondary.
Before executing the syncrvg command to perform
synchronization, use the vradmin stoprep command to stop
replication to the Secondary, sechost.
Always use the -c checkpoint option during the RVG
synchronization. When checkpoint is specified, the syncrvg
command starts the checkpoint, performs the RVG
synchronization, and stops the checkpoint when the
synchronization is complete. After the RVG synchronization
is complete, start replication by running the vradmin
startrep command with the specified checkpoint.
- 15 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
Use the -verify option to verify and report the data
differences between Primary data volumes and the
corresponding Secondary data volumes. When used with
-verify option, the syncrvg command only reports the
differences between the Primary and Secondary data volumes;
it does not synchronize the Secondary volumes with the
Primary volumes. An MD5 checksum is used to calculate the
difference between the Primary and the Secondary data
volumes.
All applications using the Primary data volumes must be
stopped (or quiesced) before running the syncrvg command
with the -verify option. At the start of the -verify
syncrvg command, you are asked to confirm that the Primary
data volumes are not in use. You can use the -s option to
skip this confirmation step.
The syncrvg command reports the progress of synchronization
or verification with the following column headings:
Eps_time [Toc] [Back]
Total time that has elapsed (hh:mm:ss).
Dest_host [Toc] [Back]
IP address of the host on which the current destination
volume resides.
Src_vol [Toc] [Back]
Source volume.
Dest_vol [Toc] [Back]
Destination volume. (Data on the destination volume is
synchronized or verified with data on the source
volume.)
F'shed/Tot_sz
Amount synchronized or verified so far, and total size
of the source volume.
Diff Percentage of differences found. This percentage is
relative to the total amount synchronized or verified
so far. For full synchronization, this field is
irrelevant and is displayed as "--".
Done Percentage of the total size synchronized or verified
so far.
syncvol Synchronizes or verifies volumelist on remote host with
volumelist on local host. volumelist is a comma-separated
list of volume names. During synchronization, data on these
local volumes is transferred to the remote host. host is
- 16 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
the resolvable name or IP address of the remote host. The
remote disk group and remote volume names are optional.
When not specified, these names default to the same names as
on the local host.
To synchronize or verify remote volumes with local volumes
having different names, remote diskgroup and volumelist must
be specified.
syncvol performs difference-based synchronization by
default. To perform full synchronization, specify the -full
option. See the syncrvg keyword for a description of sync
options.
Note: syncvol is useful for synchronizing volumes that are
not associated with an RVG.
Remote volumes can be synchronized or verified with local
volumes only if the /etc/vx/vras/.rdg file on the remote
host has a local disk group ID entry in it. Before writing
to or verifying with remote volumes, this entry is validated
as part of the syncvol command. A plus character (+) entry
in the /etc/vx/vras/.rdg file on a host allows syncvol
command from any remote host to this host to proceed.
The -verify option is used to verify and report data
differences between remote data volumes and the
corresponding local data volumes. When used with the
-verify option, the syncvol command only reports percentage
data differences between remote and local volumes; it does
not synchronize remote volumes with local volumes.
On invoking the syncvol command, you are asked to confirm
that you want to overwrite data on volumes on the remote
host with data from volumes on the local host (for
synchronization), or that the local and remote volumes are
not in use (for verification). You can use the -s option to
skip this confirmation step.
See the description of syncrvg for an explanation of the
output from this command.
takeover Takes over the Primary role in the Replicated Data Set (RDS)
rvg. The takeover command must be run from a Secondary host
in the RDS. This command works only when the Primary host
in the RDS is down or not reachable from the Secondary. The
data on the Secondary must be consistent for takeover to
work.
The takeover command is useful when the Primary host goes
down due to an unscheduled problem or due to a disaster.
- 17 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
During takeover the Secondary RVG is converted to a Primary
RVG. Volumes associated with Secondary RVG are not
accessible during takeover.
It is recommended that Secondary data volumes have Data
Change Maps (DCM). This allows the takeover command to
enable failback logging on the new Primary (or current
Secondary). Once failback logging is enabled, the fbsync
command can later be used to synchronize the data volumes on
the failed Primary with the data volumes on new Primary.
The takeover command enables failback logging by default.
The command fails if any of the data volumes on the
Secondary do not have associated DCMs.
Use the -N option to convert Secondary to Primary without
enabling failback logging. This requires either full
synchronization or difference-based synchronization of the
data volumes on failed Primary. Failback logging
synchronization is the recommended method of
synchronization. Refer to the fbsync command for more
details.
To automatically synchronize the failed Primary once it
becomes available, use the -autofb option with the takeover
command. This converts the failed Primary to the Secondary
once it comes up, and also uses failback logging to
synchronize the data volumes on the failed Primary.
Note: The -autofb option can be used only if Secondary data
volumes have DCMs.
OPTIONS [Toc] [Back]
-a Starts replication with autosync. This option is used to
attach the Primary RLINK as part of the startrep command.
-all Performs the specified operation to all the Secondary hosts
in the Replicated Data Set. This option can only be used
with the ibc command.
-autofb Automatically uses failback logging to convert and
synchronize a failed Primary as soon as the failed Primary
comes back up. This option can only be used with the
takeover command.
-c checkpoint
Starts a checkpoint, transfers the data to remote host, and
marks the checkpoint end, if specified with the syncrvg
command. This checkpoint can be used to attach the Primary
RLINK so that writes made to Primary volumes during
synchronization are transferred to Secondary volumes.
- 18 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
This option can also be used with the startrep command to
indicate the checkpoint to which the Primary RLINK needs to
be attached.
-f Forces an operation that is usually not allowed by vradmin.
When used with the delvol command, this option removes a
volume from the RDS even when the Primary RVG is not
stopped.
When used with the startrep command, this option starts
replication by attaching the Primary RLINK with the -f
option (as done by the vxrlink att command).
When used with the stoprep command, this option stops
replication to a Secondary even when the Primary RLINK is
not up-to-date.
When used with the delpri command, this option deletes a
Primary RVG even when it has not been stopped.
When used with the resizevol command, this option allows
shrinking a data volume size in an RDS.
-full Performs full synchronization of volumes on local and remote
hosts. This option works only with the syncrvg and syncvol
commands.
-g diskgroup
Specifies the disk group for the operation, either by disk
group ID or by disk group name. By default, the disk group
is chosen based on the RDS rvg (if any).
-l When used with the printrvg command, this option displays
RDS information in long format. In the long format,
additional information such as data volume count, SRL name,
and RLINK-related information are displayed for each RVG in
the RDS. This option also displays any configuration errors
in the RDS.
When used with the repstatus command, this option displays
additional replication information such as Primary and
Secondary RLINK names, replication mode configured, and
settings for latency and SRL protection for each Secondary
RVG in the RDS.
-N Allows takeover without failback logging. This option can
be used only with the takeover command.
-nodcm Skip adding DCMs to data volumes. This option can be used
with the addsec, addvol, and createpri commands.
- 19 - Formatted: January 24, 2005
vradmin(1M) VxVM 3.5 vradmin(1M)
1 Jun 2002
-s Works without prompting the user for confirmation. Specify
this option to include the addvol, migrate, stoprep,
[-verify] syncrvg and [-verify] syncvol commands in noninteractive
shell scripts.
-sdg diskgroup
Specifies the disk group name on the Secondary host for the
addsec operation.
-verify Verifies and reports percentage differences between volumes
on local and remote hosts. The volumes are not
synchronized. This option works only with syncrvg and
syncvol.
-version Displays vradmin version information.
-wait Blocks until failback synchronization is complete.
FILES [Toc] [Back]
/etc/vx/vras/.rdg File containing disk group IDs for
authentication in addsec and syncvol.
/etc/vx/vvr/ibc_scripts Directory containing scripts used in ibc
command.
/var/vx/vras/log Directory containing log files for
vradmind server.
EXIT CODES [Toc] [Back]
The vradmin utility exits with a status of 1 if the attempted
operation fails. This exit code is the same for all problems.
SEE ALSO [Toc] [Back]
vxassist(1M), vxedit(1M), vxibc(1M), vxintro(1M), vxmake(1M),
vxprint(1M),
|
