cvadmin(8) cvadmin(8)
NAME
cvadmin - Administer an Xsan File System
SYNOPSIS
cvadmin [-H FSMHostName] [-F VolumeName] [-M] [-f filename]
[-e command1 -e command2...] [-x]
DESCRIPTION
cvadmin is an interactive command used for general purpose administra-
tion of an Xsan volume including:
1. displaying file system and client status
2. activating a file system currently in stand-by mode
3. viewing and modifying storage pool attributes
4. enabling File System Manager (FSM) tracing
5. displaying disk and path information for the local system
6. forcing FSM failover
7. fetching FSM usage and performance statistics
8. temporarily enabling or disabling global file locking
9. generating a report of open files
10. listing currently held file locks
11. starting, restarting and stopping of daemon processes
12. resetting RPL information
OPTIONS
Invoke cvadmin to start the interactive session and list the running
File System Managers (FSMs). (Note: Xsan system services must be
started prior to running cvadmin. In particular, the local fsmpm(8)
process must be active.)
Then (optionally) use the select command described below to pick an FSM
to connect to. Once connected, the command will display basic informa-
tion about the selected volume and prompt for further commands.
Note that a few commands such as paths, disks, start, and stop require
information obtained from the local fsmpm(8) only, so there is is no
need to select an FSM prior to using them.
USAGE
-H FSMHostName
Connect to the FSM located on the machine FSMHostName. By de-
fault cvadmin will attempt to connect to an FSM located on the
local machine.
-F VolumeName
Automatically set the volume VolumeName as the active volume in
cvadmin.
-M When listing file systems with the select command, display [man-
aged] next to each file system with DataMigration enabled. This
option is currently only intended for use by support personnel.
-f filename
Read commands from filename
-e command
Execute command(s) and exit
-x Enable extended commands.
COMMANDS
The cvadmin commands can be used to display and modify the Xsan active
configuration. When a modification is made, it exists only as long as
the FSM is running. More permanent changes can be made in the configu-
ration file. Refer to the snfs_config(5) man page for details. The fol-
lowing commands are supported.
activate volume_name [hostname_or_IP_address]
Activate a volume volume_name. This command may cause an FSM to
activate. If the FSM is already active, no action is taken.
activate volume_name number_of_votes
Apple Internal only. Bypass the election system and attempt to
activate the fsm on this node.
clientunmount gentle_time_in_seconds
clientunmount hard gentle_time_in_seconds force_time_in_seconds
Instruct clients of the selected volume to unmount. The first
form of the command waits gentle_time_in_seconds seconds for
clients to unmount the volume. Clients receive a message from
the fsm directing them to normally unmount the volume. During
this time, no new client connections are permitted. If all
clients unmount within that time, the fsm continues with new
connections forbidden.
In the second form of the command, the fsm moves to a second
stage if any clients fail to unmount. In this stage, clients are
instructed to unmount the volume forcibly. The fsm waits
force_time_in_seconds seconds, and then completes the operation.
Client connections remain forbidden at the end of this opera-
tion.
After completing this command, the administrator should stop the
volume.
debug [[+|-] flag [ ... ]]
View or set the File System Manager's debugging flags. Entering
the command with no flag will return current settings, the loca-
tion of the FSM log file and a legend describing what each set-
ting does. By entering the command with a flag list, the FSM
Debugging Flags will be set accordingly. Each flag can be ei-
ther a name or numeric value. Names will be mapped to their nu-
meric value, and may be abbreviated as long as they remain
unique. Numeric values are specified using a standard decimal
or hexadecimal (0x) value of up to 32 bits. Using '+' or '-'
enables ('+') or disables ('-') only the selected flags, leaving
all other flags unchanged.
NOTE - Setting Debugging Flags will severely impact the FSM's
performance! Do this only when directed by an Apple specialist.
disks [refresh]
Display the Xsan disk volumes local to the system that cvadmin
is attached to. Using the optional refresh argument will force
the fsmpm to re-scan all volumes before responding. If the
fsmpm's view of the disks in any file system changes compared
with the FSM's view of that client's disks as a result of the
refresh, a disconnect and reconnect to the FSM will take place
to resynchronise the file system state.
disks [refresh] fsm
Display the Xsan meta-data disk volumes in use by the fsm. If
the optional refresh argument is used, additional paths to these
volumes may be added by the fsm.
down poolname
Down the storage pool poolname. This will down any access to the
storage pool.
fail [volume_name|index_number]
Initiate an FSM Failover of volume volume_name. This command may
cause a stand-by FSM to activate. If an FSM is already active,
the FSM will shut down. A stand-by FSM will then take over. If a
stand-by FSM is not available the primary FSM will re-activate
after failover processing is complete.
files Report counts of files, directories, symlinks and other objects
which are anchored by a user type inode. These include named
streams, block and character device files, fifos or pipes and
named sockets. If the file system is undergoing conversion to
StorNext 5.0, conversion progress is displayed and counters re-
flect the count of converted objects.
fsmlist [volume_name] [ on hostname_or_IP_address]
Display the state of FSM processes, running or not. Optionally
specify a single volume_name to display. Optionally specify the
host name or IP address of the system on which to list the FSM
processes.
filelocks
Query cluster-wide file/record lock enforcement. Currently
cluster-wide file locks are automatically used on Unix. Windows
file/record locks are optional.
If enabled, byte-range file locks are coordinated through the
FSM, allowing a lock set by one client to block overlapping
locks by other clients. If disabled, then byte-range locks are
local to a client and do not prevent other clients from getting
byte-range locks on a file, however they do prevent overlapping
lock attempts on the same client.
help (?)
The help or ? command will display a command usage summary.
latency-test [index_number|all] [seconds]
Run an I/O latency test between the FSM process and one client
or all clients. The default test duration is 2 seconds.
metadata
Report metadata usage. Also provide an estimate on the value of
bufferCacheSize that will allow all metadata to be cached.
metadump { status | rebuild | suspend | resume }
Manage the metadump functionality of the selected FSM.
The status command prints the progress of the current metadump
activity, if any. If capturing a new metadump or restoring an
existing one, the percentage complete will be displayed. Other-
wise, the current update backlog is displayed.
The rebuild command will force the FSM to discard the existing
metadump and capture a new one. This is performed online.
The suspend and resume commands are used internally to facili-
tate backups of the metadump files. They should not be invoked
manually except under direction from support.
multipath groupname {balance|cycle|rotate|static|sticky}
Xsan has the capability of utilizing multiple paths from a sys-
tem to the SAN disks.
This capability is referred to as "multi-pathing", or sometimes
"multi-HBA support". (HBA := Host Based Adaptor).
At "disk discovery" time, for each physical path (HBA), a scan
of all of the SAN disks visible to that path is initiated, accu-
mulating information such as the Xsan label, and where possible,
the disk (or LUN) serial number.
At mount time, the visible set of Xsan labeled disks is matched
against the requested disks for the volume to be mounted.
If the requested disk label appears more than once, then a "mul-
ti-path" table entry is built for each available path.
If the disk (or LUN) device is capable of returning a serial
number, then that serial number is used to further verify that
all of the paths to that Xsan labeled device share the same se-
rial number.
If the disk (or LUN) device is not capable of returning a serial
number then the device will be used, but Xsan will not be able
to discern the difference between a multi-path accessible de-
vice, and two or more unique devices that have been assigned du-
plicate Xsan labels.
The presence of serial numbers can be validated by using the
"cvlabel -ls" command. The "-s" option requests the displaying
of the serial number along with the normal label information.
There are five modes of multi-path usage which can also be spec-
ified in the filesystem config file. In cases where there are
multiple paths and an error has been detected, the algorithm
falls back to the rotate method. The balance and cycle methods
will provide the best aggregate throughput for a cluster of
hosts sharing storage.
balance
The balance mode provides load balancing across all the
available, active, paths to a device. At I/O submission
time, the least used HBA/controller port combination is
used as the preferred path. All Xsan File System I/O in
progress at the time is taken into account.
cycle The cycle mode rotates I/O to a LUN across all the avail-
able, active, paths to it. As each new I/O is submitted,
the next path is selected.
rotate The rotate mode is the default for configurations where
the operating system presents multiple paths to a device.
In this mode, as an I/O is initiated, an HBA controller
pair to use for this I/O is selected based on a load bal-
ance method calculation.
If an I/O terminates in error, a "time penalty" is as-
sessed against that path, and another "Active" path is
used. If there are not any "Active" paths that are not
already in the "error penalty" state, then a search for
an available "Passive" path will occur, possibly trigger-
ing an Automatic Volume Transfer to occur in the Raid
Controller.
static The "default" mode for all disks other than Dual Raid
controller configurations that are operating in Ac-
tive/Active mode with AVT enabled.
As disks (or LUNs) are recognized at mount time, they are
statically associated with an HBA in rotation.
i.e. given 2 HBA's, and for disks/LUNs:
disk 0 -> HBA 0
disk 1 -> HBA 1
disk 2 -> HBA 0
disk 3 -> HBA 1
and so on...
sticky In this mode, the path to use for an I/O is based on the
identity of the target file. This mode will better uti-
lize the controller cache, but will not take advantage of
multiple paths for a single file.
The current mode employed by a storage pool can be viewed via
the "cvadmin" command "show long", and modified via the "cvad-
min" command "multipath".
Permanent modifications may be made by incorporating a "Multi-
PathMethod" configuration statement in the configuration file
for a storage pool.
In the case of an I/O error, that HBA is assessed an "error
penalty", and will not be used for a period of time, after which
another attempt to use it will occur.
The first "hard" failure of an HBA often results in a fairly
long time-out period (anywhere from 30 seconds to a couple of
minutes).
With most HBA's, once a "hard" failure (e.g. unplugged cable)
has been recognized, the HBA immediately returns failure status
without a time-out, minimizing the impact of attempting to re-
use the HBA periodically after a failure. If the link is re-
stored, most HBA's will return to operational state on the next
I/O request.
paths Display the Xsan disk volumes visible to the local system. The
display is grouped by <controller> identity, and will indicate
the "Active" or "Passive" nature of the path if the Raid Con-
troller has been recognized as configured in Active/Active mode
with AVAT enabled.
proxy [long]
Display Disk Proxy servers and optionally display the disks they
serve for this volume.
proxy who hostname
The "who" option displays all proxy connections for the speci-
fied host.
qos Display per-stripe group QOS statistics. Per-client QoS statis-
tics are also displayed under each qos-configured stripe group.
quit This command will disconnect cvadmin from the FSM and exit.
ras enq event "detail string"
Generate an SNFS RAS event. For internal use only.
ras enq event reporting_FRU violating_FRU "detail string"
Generate a generic RAS event. For internal use only.
repfl Generate a report that displays the file locks currently held.
Note: this command is only intended for debugging purposes by
support personnel. In future releases, the format of the report
may change or the command may be removed entirely. Running the
repfl command will write out a report file and display the out-
put filename.
repof Generate a report that displays all files that are currently
open on each Xsan client. Only file inode numbers and stat in-
formation are displayed, filenames are not displayed. Running
the repof command will write out a report file and display the
output filename. In future releases, the format of the report
may change.
resetrpl [clear]
Repopulate Reverse Path Lookup (RPL) information. The optional
clear argument causes existing RPL data to be cleared before
starting repopulation. Note: resetrpl is only available when
cvadmin is invoked with the -x option. Running resetrpl may
significantly delay FSM activation. This command is not intend-
ed for general use. Only run resetrpl when recommended by Tech-
nical Support.
restartd daemon [once]
Restart the daemon process. For internal use only.
select [volume_name|N]
Select an active FSM to view and modify. If no argument is
specified, a numbered list of FSMs and running utilities will be
displayed. If there is only one active volume in the list, it
will automatically be selected.
When a running utility is displayed by the select command, it
will show the following information. First the name of the file
system is displayed. Following that, in brackets "[]", is the
name of the utility that is running. Third, a letter indicating
the access type of the operation. The options here are (W) for
read-write access, (R) for read-only access and (U) for unique
access. Finally, the location and process id of the running
utility is displayed.
If volume_name is specified, then cvadmin will connect to the
current active FSM for that volume. If N (a number) is speci-
fied, cvadmin will connect to the Nth FSM in the list. However,
only active FSMs may be selected in this form.
show [poolname] [long]
Display information about the storage pools associated with the
selected volume. If a storage pool name poolname is given only
that storage pool's information will be given. Omitting the
poolname argument will display all storage pools associated with
the active file system. Using the long modifier will additional-
ly display detailed information about the disk units associated
with displayed storage pools.
start volume_name [on hostname_or_IP_address]
Start a File System Manager for the volume volume_name. When
the command is running on an MDC of an HA cluster, the local FSM
is started, and then an attempt is made to start the FSM on the
peer MDC as identified by the /Library/Preferences/Xsan/ha_peer
file. When the optional hostname_or_IP_address is specified,
the FSM is started on that MDC only. The volume's configuration
file must be operational and placed in /Library/Prefer-
ences/Xsan/<volume_name>.cfg before invoking this command. See
snfs_config(5) for information on how to create a configuration
file for an Xsan volume.
startd daemon [once]
Start the daemon process. For internal use only.
stat Display the general status of the volume. The output will show
the number of clients connected to the volume. This count in-
cludes any administrative programs, such as cvadmin. Also shown
are some of the static file-system-wide values such as the block
size, number of storage pools, number of mirrored storage pools
and number of disk devices. The output also shows total blocks
and free blocks for the entire volume.
stats client_IP_address [clear]
Display read/write statistics for the selected file system.
This command connects to the host FSMPM who then collects sta-
tistics from the file system client. The ten most active files
by bytes read and written and by the number of read/write re-
quests are displayed. If clear is specified, zero the stats af-
ter printing.
stop volume_name [on hostname_or_IP_address]
Stop the File System Manager for volume_name. This will shut
down the FSM for the specified volume on every MDC. When the
optional hostname or IP address is specified, the FSM is stopped
on that MDC only. Further operations to the volume will be
blocked in clients until an FSM for the volume is activated.
stopd daemon
Start the daemon process. For internal use only.
up poolname
Up the storage pool poolname. This will restore access to the
storage pool.
who Query client list for the active volume. The output will show
the following information for each client.
SNFS I.D. - Client identifier
Type - Type of connection. The client types are:
FSM - File System Manager(FSM) process
ADM - Administrative(cvadmin) connection
CLI - Volume client connection. May be followed by a CLI
type character:
S - Disk Proxy Server
C - Disk Proxy Client
H - Disk Proxy Hybrid Client. This is a client that has
been configured as a proxy client but is operating
as a SAN client.
Location - The clients hostname or IP address
Up Time - The time since the client connection was initiated
License Expires - The date that the current client license will
expire
EXAMPLES
Invoke the cvadmin command for FSM host cornice, volume named default.
spaceghost% cvadmin -H k4 -F xsan1
Xsan File System Administrator
Enter command(s)
For command help, enter "help" or "?".
List FSS
File System Services (* indicates service is in control of FS):
1>*xsan1[0] located on k4:32823 (pid 3988)
Select FSM "xsan1"
Created : Fri Jul 25 16:41:44 2003
Active Connections: 3
Fs Block Size : 4K
Msg Buffer Size : 4K
Disk Devices : 1
Stripe Groups : 1
Mirror Groups : 0
Fs Blocks : 8959424 (34.18 GB)
Fs Blocks Free : 8952568 (34.15 GB) (99%)
Show all the storage pools in the volume;
snadmin (xsan1) > show
Show stripe group(s) (File System "xsan1")
Stripe Group 0 [StripeGroup1] Status:Up,MetaData,Journal
Total Blocks:8959424 (34.18 GB) Free:8952568 (34.15 GB) (99%)
MultiPath Method:Rotate
Primary Stripe 0 [StripeGroup1] Read:Enabled Write:Enabled
Display the long version of the RegularFiles storage pool;
snadmin (xsan1) > show StripeGroup1 long
Show stripe group "StripeGroup1" (File System "xsan1")
Stripe Group 0 [StripeGroup1] Status:Up,MetaData,Journal
Total Blocks:8959424 (34.18 GB) Free:8952568 (34.15 GB) (99%)
MultiPath Method:Rotate
Stripe Depth:1 Stripe Breadth:16 blocks (64.00 KB)
Affinity Set:
Realtime limit IO/sec:0 (~0 mb/sec) Non-Realtime reserve IO/sec:0
Committed RTIO/sec:0 Non-RTIO clients:0 Non-RTIO hint IO/sec:0
Disk stripes:
Primary Stripe 0 [StripeGroup1] Read:Enabled Write:Enabled
Node 0 [disk002]
Down the storage pool named stripe1;
snadmin (xsan1) > down stripe1
Down Stripe Group "stripe1" (File System "xsan1")
Stripe Group 0 [stripe1] Status:Down,MetaData,Journal
Total Blocks:2222592 (8682 Mb) Free:2221144 (8676 Mb) (99%)
Mirrored Stripes:1 Read Method:Sticky
Primary Stripe 0 [stripe1] Read:Enabled Write:Enabled
Disable reads on the mirrored storage pool stripe1m.
snadmin (xsan1) > disable stripe1m read
Disable Stripe Group "stripe1m" (File System "xsan1")
Stripe Group 0 [stripe1] Status:Down,MetaData,Journal
Total Blocks:2222592 (8682 Mb) Free:2221144 (8676 Mb) (99%)
Mirrored Stripes:1 Read Method:Sticky
Primary Stripe 0 [stripe1] Read:Enabled Write:Enabled
FILES
/Library/Preferences/Xsan/*.cfg
SEE ALSO
cvfs(8), snfs_config(5), fsmpm(8), fsm(8), mount_acfs(8)
Xsan File System August 2015 cvadmin(8)
Mac OS X 10.12.3 - Generated Thu Feb 9 18:11:47 CST 2017