Applies To:

Show Versions Show Versions

Manual Chapter: Managed-Volume Troubleshooting Tools
Manual Chapter
Table of Contents   |   << Previous Chapter   |   Next Chapter >>

40 
The cancel migration command stops a file migration that is in progress. You can cancel all migrations in a managed volume or the migration of a single file.
cancel migration ns volume vol-path
ns (1-30 characters) is the namespace where a migration is in process.
vol-path (1-1024 characters) specifies a managed volume.
file-name (1-1024 characters) is the virtual path of a file that is being migrated. This includes the path name for the managed volume itself (for example, the file name for mydir/myfile.doc in the /bigvol volume would be /bigvol/mydir/myfile.doc).
bstnA# cancel migration medarcv volume /rcrds
bstnA# cancel migration file /claims/2007/masterlist.xsl
The cancel remove command stops a share-removal that is in progress.
cancel remove namespace ns volume vol-path share share-name
ns (1-30 characters) is the namespace from which the share is being removed.
vol-path (1-1024 characters) specifies the namespace volume.
share-name (1-64 characters) is the share that is being removed.
You can remove a share with remove-share migrate, remove-share nomigrate, no share, or no filer. The first stage of the share removal is a scan of the filer; this reads all file attributes on the shares back-end directories. During this phase (the longest in the process), you can cancel the share removal. After this phase is over, the CLI does not permit anyone to cancel the operation.
bstnA# cancel remove namespace ns volume /vol share testrun
clear statistics filer ext-filer-name [share share-name]
ext-filer-name (optional, 1-64 characters) identifies the external filer by its configured name. Use show external-filer to display all configured external filers. If you omit this argument, the command clears all traffic statistics on the system. This option is not recommended, as it results in incomplete statistics for some forms of show statistics namespace.
share-name (optional, 1-64 characters) clears traffic statistics for this specific back-end share. This is the name of the share at the back-end filer, not at the namespace. As above, this option is not recommended; it results in incomplete statistics for some forms of show statistics namespace.
The show statistics filer command shows the NFS and/or CIFS traffic between the ARXs software and its back-end filers. Use this command to clear those traffic counters. This also clears the traffic counters for the various forms of show statistics namespace: show statistics namespace ... fastpath, show statistics namespace ... request-detail, show statistics namespace ... response-detail, and show statistics namespace ... summary.
bstnA# clear statistics filer das1
Managed volumes use metadata to track the locations of their files on back-end storage. They store the metadata on an external metadata share, and keep statistics on the usage of that share. Use this command to clear these statistics for the entire system, a specific namespace, or particular managed volume.
clear statistics metadata [namespace ns [volume vol]]
ns (optional, 1-30 characters) identifies a namespace.
vol (optional, 1-1024 characters) is a managed volume.
The show statistics metadata command shows usage statistics for the metadata shares behind your managed volumes. Use this command to clear these statistical counters for the whole system, one namespace, or one managed volume.
bstnA# clear statistics metadata
Volume software stores metalog data to be used for recovery from failovers (see redundancy) and crashes. The volume software needs to read and write this metalog data quickly to ensure proper performance; you can use the show statistics metalog command to monitor the read/write statistics for metalog storage. Use this command to clear these statistics for the entire system or a specific volume-group.
clear statistics metalog [volume-group vol-group-id]
vol-group-id (optional, 1-255 characters) identifies a volume group. You can use the show volume-group command to show all volume groups, which volumes are in each of them, and their ID numbers.
The show statistics metalog command shows usage statistics for the metalog storage behind your managed volumes. These statistics record the latency for metalog reads and writes. Use this command to clear these statistical counters for the whole system or a single volume group.
bstnA# clear statistics metalog volume-group 9
clear statistics migration namespace [volume]
namespace (1-30 characters) identifies a namespace.
volume (optional, 1-1024 characters) is a managed volume.
The show statistics migration command shows file-migration statistics for managed volumes. Use this command to clear the statistical counters for one namespace or managed volume.
bstnA# clear statistics migration medarcv /rcrds
export-mapping [outputfile outputfile] [nfs | cifs]
outputfile outputfile (optional; 5-128 characters) specifies a prefix for a customized file name (for example, jrandom_exp_map.rpt), as opposed to the default. Use the show reports command to display the file in the maintenance directory.
nfs | cifs (optional) narrows the report to a single protocol, NFS or CIFS.
outputfile outputfile - export-mapping.rpt.
The name of the report appears after you issue the command. For long reports, you can use tail to follow the report as it is written. Use show reports export-mapping.rpt to read the report. You can search through the report with grep. To copy or delete it, use the copy or delete commands. If you want to truncate the report before it finishes, use the truncate-report command.
The report is divided into three tables. The first shows all front-end services, the next shows all back-end filers, and the third table maps all front-end shares to a single back-end path. The single back-end path is the master directory for the root of the front-end share; by default, all of the front-end shares subdirectories and files go to this back-end path. If it is the only share in the volume, or if policy is not running in the volume, this contains all of the files and directories in the front-end share. Otherwise, files and directories may be distributed among the volumes shares; use show server-mapping for a map of all back-end shares behind each front-end share.
bstnA# export-mapping
bstnA# export-mapping outputfile nfs_only.rpt nfs
bstnA# show reports export-mapping.rpt
bstnA# show reports nfs-only.rpt
Use the find command to locate one files back-end filer and path.
find host hostname-or-ip {nfs| cifs} share-name path path [verbose]
hostname-or-ip identifies a virtual server by its external DNS hostname or Virtual-IP address. Use show virtual service for a list of all virtual servers and their Virtual IP addresses. (To look up DNS names at an external DNS server, you must first identify the server with the ip name-server command.)
nfs | cifs is a required choice, determining the type of share that follows.
share-name (1-4096 characters) is the name of the front-end share, as seen by clients. The show server-mapping command shows all front-end shares on the ARX, in the left column.
path (1-4096 characters) is the virtual-file path within the front-end share.
verbose (optional) adds the files NFS filehandles to the output.
find global-server fqdn {nfs| cifs} share-name path path [verbose]
fqdn (1-255 characters) identifies a global server by its fully-qualified domain name (FQDN). Use show global service for a list of all global servers and their FQDNs.
find namespace namespace path path [verbose]
namespace (1-30 characters) identifies the namespace. Use show namespace for a list of all namespaces.
path (1-4096 characters) is the virtual-file path within the namespace. This is the file path as seen by the client, starting with the volume path.
verbose (optional) adds the files NFS filehandles to the output.
find wins wins-name {nfs| cifs} share-name path path
wins-name (1-255 characters) identifies a virtual server by its optional wins-name or one of its optional wins-alias names.
NFS Physical Location is an NFS path to the physical file. This appears in ip-address:/path-to-file format.
Managed Volume Path indicates that the path is on a direct volume, and the physical location is on an ARX managed-volume that is standing in as a filer.
CIFS Physical Location is a CIFS path to the physical file. This appears in //ip-address/path-to-file format. If this is an NFS-only file in a multi-protocol (CIFS and NFS) volume, a message here indicates that the file has an inconsistent name. You can use the nsck ... report inconsistencies command to find all of the volumes inconsistent names.
The verbose output shows two NFS filehandles for the file:
Virtual File Handle is the filehandle that the ARX presents to clients.
Physical File Handle is the filehandle that the server gave to the ARX.
The filehandles may be incomplete for the find namespace command. Use one of the other versions (such as find host) for complete NFS filehandles.
The find command identifies a files location now; for a files location in the past, you can use file tracking. File tracking can be useful for finding the correct filer-backup tape for a lost or compromised file.
The file-tracking feature requires some configuration before you can make any file queries. Specifically, a managed volume requires a snapshot rule to regularly copy its configuration and metadata to a file-history archive. After some copies of metadata have been archived, you can use the show file-history virtual-service command to query file locations at different dates.
bstnA# find namespace wwmed path /acct/index.html
bstnA# find host 192.168.25.14 nfs /claims path /stats/piechart.ppt
bstnA# find global-server insur.medarch.org nfs /claims path /stats/piechart.ppt
bstnA# find global-server insur.medarch.org cifs CLAIMS path /index.html verbose
Use the remove service command to remove a namespace and all other configuration objects that are exclusively dedicated to the namespace (such as global servers, proxy users, and/or NFS access-lists).
remove service namespace [timeout seconds] [sync]
namespace (1-30 characters) identifies the namespace.
seconds (optional, 300-36,000) is a time limit for each component operation in this service removal, in seconds.
sync (optional) shows the operations progress at the command line. With this option, the CLI prompt does not return until all components have been removed.
By default, this command generates a report to show all of the actions it takes to remove the volume(s), in order. The CLI shows the report name after you issue the command, and then returns. You can enter CLI commands as the namespace software removes the objects in the background. Use tail to follow the report as it is written. Use show reports file-name to read the report. You can search through the report with grep. To copy or delete it, use the copy or delete commands. Use the sync option to send the status to the command line instead; the command does not generate a report if you use the sync option.
If you remove a cifs service with any dynamic-dns names, you must remove all of the Dynamic DNS names (with no dynamic-dns) before you run this command.
Use remove namespace to remove a namespace or volume without affecting any global servers, NFS-access-control lists, or external filers. To remove only the policy objects from the namespace or a single volume, use remove namespace ... policy-only. The remove namespace ... volume ... exports-only command finds all of a volumes front-end exports and removes them. To remove a share from a volume, use remove-share migrate or remove-share nomigrate.
bstnA# remove service medco
Use the remove-share migrate command to remove a share from a namespace volume and move all of its files to another share or share farm in the same volume. The volumes clients are unaffected by this share removal.
remove-share migrate namespace volume share dest
[close-file [exclude fileset]] [async]
namespace (1-30 characters) identifies the namespace.
volume (1-1024 characters) is the name of the volume.
share (1-64 characters) is share being removed.
dest (1-64 characters) is the new destination share or share farm for files and master directories.
close-file (optional) causes the operation to close all files open by CIFS clients in order to migrate them. If you omit this option, files opened by CIFS clients cannot migrate off of the selected share. This would leave files stranded on the share and cause the share removal to fail.
exclude fileset (optional if you choose close-file, 1-64 characters) is a fileset to exclude from automatic closure. If a file in this fileset is open through CIFS, the rule places it on a retry queue instead of automatically closing it. If such a file remains open for the duration of the share removal, it never migrates off of the share and the share-removal fails.
async (optional if you specify a dest) makes this command return immediately, rather than waiting for the share removal to finish.
This command does not apply to direct volumes (see direct); only metadata-based managed volumes.
This is not recommended in a volume that supports snapshots. In those volumes, you should use a place-rule to drain the share first, wait until all retained snapshots have aged out, and then run this command or remove-share nomigrate.
Clients are not impacted by the remove-share command. All files that migrate to the new back-end share remain in the same volume and namespace, so their file associations remain intact. Front-end clients see no change.
A volume uses the snapshots on a replica-snap share and ignores all other files and directories. This type of share does not have any files or directories to migrate, so the dest share, close-file, and exclude options are all ignored.
drain_share_rule_for_share_share-name_time.rpt, and
remove.job-id.share-name.share-id.rpt
Use show reports to view either of these reports.
prtlndA# remove-share migrate nemed /vol44 rxbills globalmedco close-file
Use the remove-share nomigrate command to remove a share from a namespace without migrating any of its files. This causes all of the shares files (if there are any) to be removed from the volume.
remove-share nomigrate namespace volume share [dest [async]]
namespace (1 - 30 characters) is the name of the namespace.
volume (1 - 1024 characters) is the name of the volume.
share (1 - 64 characters) is the name of the share.
dest (not allowed for replica-snap shares, but required for all other share types; 1 - 64 characters) is the new destination share or share farm where the original shares primary copies of directories (if identified) are to be migrated. Use show global-config namespace for a list of shares and share farms in the namespace.
async (optional if you specify a dest) makes this command return immediately, rather than waiting for the share removal to finish.
This command does not apply to direct volumes (see direct); only metadata-based managed volumes.
If the share contains any master directories in the volume (that is, the first-imported instance of a directory), they are copied to the dest share or share farm. This may fail if the remove process encounters any of the following issues:
A master directory may not be able to migrate to a dest that supports CIFS if the destination filer is very low on free space. The CIFS ACL(s) from the master directory could exceed the free space on the destination. You can resolve this issue by choosing a dest with more free space behind it, increasing disk space on the destination filer(s), or using a place-rule (with a source share) to migrate files from the dest share to an emptier share in the same managed volume.
A volume uses the snapshots on a replica-snap share and ignores all other files and directories. This type of share does not have any master directories to migrate, so a dest share is not required.
This command creates a report, remove.job-id.share-name.share-id.rpt. Use show reports to view its contents.
bstnA# remove-share nomigrate nemed /vol33 oldScripts globalmedco
Use the remove-share offline command to remove an offline, unreachable, or otherwise defunct back-end share from a namespace.
remove-share offline namespace volume share dest [async]
namespace (1 - 30 characters) is the name of the namespace.
volume (1 - 1024 characters) is the name of the volume.
share (1 - 64 characters) is the name of the share.
dest (1 - 64 characters) is the new destination share or share farm where the original shares master directories (if any) are to be migrated. A master directory is the first-imported instance of a directory in the managed volume. Use show global-config namespace for a list of shares and share farms in the namespace.
async (optional) makes this command return immediately, rather than waiting for the share removal to finish.
Important: This operation migrates any master directories from the defunct share to the dest. Because the share is unreachable, this operation cannot probe those directories to find their file attributes and/or ACLs. The file attributes for the master directories are therefore set to zero at the dest.
This command does not apply to direct volumes (see direct); only metadata-based managed volumes.
This command creates a report, remove.job-id.share-name.share-id.rpt. Use show reports to view its contents.
As mentioned above, this command migrates all master directories with 0 (zero) file attributes. Look in the remove report to find these directories: in the section below Prepare Start Time, look for [AR] next to each affected directory. You can access the volume through its VIP to manually reset the attributes of these directories.
bstnA# remove-share offline nemed /vol33 rxpresc globalmedco
Use the show policy history command to view the policy events that have occurred in a volume, or the detailed events for a specific rule or share farm.
show policy history namespace namespace volume volume [rule rule-name]
namespace (1-30 characters) is the namespace.
volume (1-1024 characters) identifies the volume.
rule-name (optional, 1-1024 characters) narrows the scope to one rule or share farm in the volume. This shows detailed history for the rule or share farm.
The show policy command shows all namespace policies. Use show policy filesets to show all filesets, and use show schedule to show all schedules. The show policy queue command shows all files currently waiting to be migrated, if any.
bstnA# show policy history namespace insur volume /claims
prtlndA# show policy history namespace nemed volume /acctShdw rule farm1
bstnA# show policy history namespace insur volume /claims
prtlndA# show policy history namespace nemed volume /acctShdw rule farm1
Use the show policy queue command to see which files (if any) are currently queued for migration.
show policy queue namespace namespace volume volume [auto-close]
namespace (1-30 characters) is the namespace.
volume (1-1024 characters) identifies the volume.
auto-close (optional) limits the output to files that have been auto-closed. (A file-placement rule can auto-close any file that is already opened through CIFS, and can hold it closed until the file is migrated. The migrate close-file command enables this feature.)
File, on the top line, is the file being migrated. This is shown from the clients perspective.
Rule is the name of the rule or share farm that is migrating the file.
Target is the name of the target share or share farm. This is the name of the share from the namespace configuration; use the show global-config namespace command to show all shares in a namespace.
Status is the third line. This explains the current status of the file migration.
If the queue is very full because migrations are repeatedly failing, you can use the policy migrate-delay command to stop the policy engine from infinitely retrying its migrations. This causes the policy engine to remove all old entries from the queue. You can also use policy migrate-retry-delay to change the time between migration retries.
The cancel migration command clears the queue.
bstnA# show policy queue namespace medarcv volume /rcrds
bstnA# show policy queue namespace medarcv volume /rcrds
Use the show statistics filer command to show the numbers of NFS RPCs and CIFS SMB commands from various software subsystems to one or more filers.
show statistics filer ext-filer-name [share {shr | all}] [nfs | cifs] [summary | request-detail | response-detail]
nfs | cifs (optional) limits the output to NFS statistics or CIFS statistics.
summary | request-detail | response-detail is an optional choice, where
summary shows the counts of each NFS and/or CIFS request, the counts of the corresponding responses, and the average round-trip times (RTTs). This is the default output if you select none of these options.
request-detail shows the counts of each NFS and/or CIFS request from each software component.
response-detail shows average RTT for each NFS and/or CIFS response to each software component.
ext-filer-name (optional, 1-64 characters) identifies the external filer by its configured name. Use show external-filer to display all configured external filers.
shr | all is a required choice if you use the optional share keyword:
shr (1-64 characters) focuses on traffic to this specific back-end share. This is the name of the share at the back-end filer, not at the namespace.
all shows all shares on the filer. This creates a separate table for each share; if you omit the share all option, the output contains a single table showing the sums of all shares.
The output from the simplest syntax, show statistics filer, shows one or two tables of statistics per external-filer: a table of NFS-traffic statistics and/or a table for CIFS traffic. Each row of the NFS table shows the total packets for a particular NFS RPC. Each row in the CIFS table shows the total packets for a given CIFS SMB command. Each row also displays the average RTT for the RPC or SMB.
Filer is the name of the external filer. As mentioned above, you can use the show external-filer filer-name command to see the IP address and/or SPN for filer-name.
All shares in filer appears if you did not use the share keyword to specify a single share. This indicates that the statistics are at the filer level; they are the sum of the statistics for all of the filers imported shares. If you select a particular share, the following fields appear instead:
CIFS Share is the name of the export or share at the filer. Both names appear if this is used as a multi-protocol share.
Reset... shows the last time these statistics were cleared. The software resets these statistics whenever someone uses the clear statistics filer command or reloads the chassis.
The request-detail flag focuses on request counts. This option breaks the request-count statistics down into the software subsystems that sent them:
Data Plane is fastpath traffic through the network software. These packets are not handled by any volume-group software on the Control Plane.
Ctrl Plane (short for Control Plane) is traffic that passes through the volume-group software for further processing.
Policy shows the traffic that is initiated by the policy engine, such as migrations and shadow copies.
Probe is the statistics from the health-checking subsystem in the namespace software. This subsystem sends periodic health-check probes to every back-end share that is connected to an ARX volume.
The response-detail flag focuses on average RTTs. This option also breaks the statistics down into the software subsystems that sent the RPCs and/or SMBs.
FormatError is the number of malformed NFS packets received from the filer(s).
NetworkError counts the send and receive failures.
ResourceError is the number of times that the ARX software ran out of resources. Contact F5 Support if you see these errors.
RPCError counts any errors from the RPC layer.
TimeoutError is the number of timeouts waiting for an NFS response.
FormatError is the number of malformed SMB (CIFS) packets received from the filer(s).
NetworkError counts the send and receive failures for SMB traffic.
ResourceError shows the number of times that the ARX software had insufficient resources to complete the transaction. Contact F5 Support if you see these errors.
SignatureError counts any SMB packets with missing or incorrect signatures. You can use the cifs filer-signatures command to change the SMB-signing policy for the current namespace.
SMBError is the number of SMB packets that failed with errors that are not directly related to any of the above errors.
TimeoutError is the number of timeouts waiting for an SMB response.
stkbrgA(cfg)# show statistics filer
canbyA(cfg)# show statistics filer nfs reponse-detail
stkbrgA(cfg)# show statistics filer
canbyA(cfg)# show statistics filer nfs response-detail
Each managed volume stores metadata about the files that it keeps on its back-end shares. This metadata is stored on an external metadata share. The latency between the managed-volume software and this external metadata share must be small or the volumes performance suffers. Use the show statistics metadata command to measure the latency between namespace processes and their metadata storage.
name (1-30 characters) specifies one namespace. You can use show namespace to get a list of all namespaces.
vol (optional; 1-1024 characters) specifies one managed volume.
Volume, and
Location identify a specific metadata share. The name of the back-end filer is its external-filer name as defined on the ARX; you can use show external-filer to find its IP address, SPN, and configuration information.
Last Reset Time shows the last time the volume software started or someone cleared these statistics with clear statistics metadata.
Read Operations counts the number of reads since the above last-reset time.
Read Response Time is the average response time for the read operations,
Read Bytes is the total bytes since the Last Reset Time,
Read Rate shows the running average in Megabytes per second,
Read Errors counts the total read errors since the volume group started,
Read Operations in Progress is the number of read operations currently underway, and
Last Read Operation is a date stamp for the most-recent read operation.
bstnA# show statistics metadata
bstnA> show statistics metadata
Namespace software records important information in its metalog, possibly to be used later for a failure recovery. To ensure proper volume performance, this latency should be low. Use the show statistics metalog command to measure the latency between namespace processes and their metalog storage.
id (optional, 1-255) identifies a particular volume group. If you omit the number, this command displays all volume groups.
To ensure proper volume performance, this latency and the latency to the backup peer (measurable with show redundancy metalog) should be as low as possible. Use this command to measure the performance of namespace software with respect to metalog storage. The output contains one table per volume group, with each of the following fields:
Last Reset Time shows the last time the volume-group software started, or the last time someone used clear statistics metalog on the volume group.
Read Operations counts the number of reads since the above time.
Read Response Time is the average response time for the read operations,
Read Bytes is the total bytes read since the Last Reset Time,
Read Rate shows the running average round-trip time for read operations, in Megabytes per second,
Read Errors counts the total read errors since the volume group started,
Read Operations in Progress is the number of read operations currently underway, and
Last Read Operation is a date stamp for the most-recent read operation.
The show redundancy metalog command shows the latency for resilvering (copying) metalog data to the backup peer. If that latency is too high, slower volume processing may be noticeable by volume clients.
bstnA# show statistics metalog
bstnA# show statistics metalog volume-group 1
bstnA> show statistics metalog
bstnA> show statistics metalog volume-group 1
Use the show statistics migration command to show the file-migration activity on the current ARX.
ns (optional; 1-30 characters) specifies one namespace.
vol (optional; 1-1024 characters) specifies one managed volume.
activity (optional) shows all migrations that are currently in progress for each of the chosen volumes.
history (optional) shows the last 100 migrations that occurred in each of the chosen volumes.
The policy engine enforces a managed volumes rules by migrating files from one back-end filer to another. A place-rule causes file migrations, as does the auto-migrate command in a share-farm. Use this command to view in-flight migrations, migration history, and some migration-related statistics.
Namespace identifies a namespace.
Volume Group is one volume-group in the above namespace.
Volume shows one volume in the above volume group.
Below each volume are the statistics requested, or configuration data if you did not select activity or history. A volume without any policy associated with it does not have anything below its name.
The summary view contains a table of namespace-migration statistics followed by a table of per-volume statistics.
Last Reset Time is the last time (in UTC, not local time) that someone ran the clear statistics migration command or rebooted the chassis.
Files Migrated is the total files migrated in the namespace.
Failed Migrations shows how many migrations failed.
Migration Time is the time required to migrate all of the Files Migrated.
Data Migrated counts the total size of files successfully migrated.
Active Migration Count is the number of migrations that are currently underway. You can see the details of these migrations by re-running this command with the activity keyword.
Average Data Rate is the average rate of transfer for all migrations behind the namespace.
Average File Size are high, low, and average samplings of the files migrated in the namespace.
If you use the activity option, the output contains a Current Migration Activity table for each volume instead of the migration statistics. This table lists the files that the policy engine is currently migrating, one file per row.
The history option shows two tables per volume with the most-recent migrations (up to 100 total):
Migration History Failure Activity shows each failed migration, one per row. Each row contains four fields.
Timestamp shows when the migration failed (in UTC, not local time),
Source is the back-end filer and path where the file started.
Destination is the back-end filer and path where the policy engine attempted (and failed) to migrate the file.
Reason is the reason for the failure.
Migration History Success Activity lists all successful migrations. Each row contains the following fields.
Timestamp shows when the migration occurred (in UTC, not local time),
Source is the back-end filer and path where the file started.
Destination is the back-end filer and path where the policy engine migrated the file.
bstnA(cfg)# show statistics migration
bstnA(cfg)# show statistics migration history namespace medarcv
bstnA(cfg)# show statistics migration
bstnA(cfg)# show statistics migration history namespace medarcv
Use the show statistics namespace ... fastpath command to show raw read/write statistics between a namespace and its back-end shares.
show statistics namespace ns [volume vol [share {shr | all}]] fastpath
ns (1-30 characters) identifies the namespace.
vol (optional; 1-1024 characters) specifies one volume in the namespace.
shr | all is a required choice if you use the optional share keyword:
shr (1-64 characters) specifies one share in the volume.
all shows all shares in the volume. This creates a separate table for each share; if you omit the share all option, the output contains a single table showing the sums of all shares.
fastpath specifies that you want the raw statistics for the chosen volumes, as opposed to NFS and/or CIFS statistics (see show statistics filer). These statistics are kept by the fastpath processes on the NSM (or Data Plane).
This command focuses on the fastpath, or data-plane, communication with back-end filers. These are raw byte counts for reads and writes, and the average packet latency. To see the counts of various CIFS and/or NFS requests to the namespaces filers, you can use the show statistics namespace ... request-detail command. To see the round-trip times (RTTs) for all of these requests, you can use show statistics namespace ... response-detail. For a summary, showing request and response counts together with RTTs, use show statistics namespace ... summary. The output of all three commands includes the counts of various errors. As with this command, you can see these statistics summarized for an entire namespace, focused on a single volume, or focused on one share.
Volume identify the volume.
All shares in volume appears if you did not use the share keyword to specify a single share. This indicates that the statistics are at a volume level; they are the sum of the fastpath (data-plane) statistics for all of the volumes shares. If you select a particular share, the following fields appear instead:
Share identifies the share, as it is named in the ARX configuration. The show global-config namespace name command shows the share names in the ARX configuration.
Filer is the name of the external filer that hosts the share. Again, this is the name of the filer in the ARX configuration; you can use the show external-filer filer-name command to see the IP address and/or SPN for filer-name.
CIFS Share is the name of the export or share at the filer. Both names appear if this is used as a multi-protocol share.
Reset... shows the last time these statistics were cleared. The software resets these statistics whenever someone reloads the chassis or uses clear statistics filer on the share.
Write Calls are all measured since the share was last enabled. Use the enable (gbl-ns-vol-shr) command to enable a namespace share.
Other Calls is the sum of other Remote Procedure Calls (RPCs) including statistics operations and file opens.
All Calls is the sum of reads, writes, and other RPCs.
Average Latency is the average packet-round-trip time between the ARX and the share.
You can use the clear statistics filer command to clear the statistics from all variations of show statistics namespace. If you use that command to clear the statistics from one share in a volume but not another, the volume-level statistics in this command are inconsistent; one share may have statistics for 2 months and the other share may have statistics for only 5 minutes. To confirm that the volume has consistent statistics, run this command on each of the volumes shares and verify that the Reset... field has the same date and time for each of them.
bstnA> show statistics namespace wwmed fastpath
bstnA> show statistics namespace medarcv volume /rcrds fastpath
bstnA> show statistics namespace wwmed fastpath
bstnA> show statistics namespace medarcv volume /rcrds fastpath
Use the show statistics namespace ... request-detail command to show request counts between a namespace and its back-end shares. The request counts are broken down by request type: the output shows separate counters for each CIFS (SMB) procedure call and/or NFS RPC.
show statistics namespace ns [volume vol [share {shr | all}]] request-detail
ns (1-30 characters) identifies the namespace.
vol (optional; 1-1024 characters) specifies one volume in the namespace.
shr | all is a required choice if you use the optional share keyword:
shr (1-64 characters) specifies one share in the volume.
all shows all shares in the volume. This creates a separate table for each share; if you omit the share all option, the output contains a single table showing the sums of all shares.
request-detail causes the output to focus on request counts from ARX software to its back-end shares. This also shows error counts.
This command counts the requests from the chosen namespace, volume or share to its back-end filers. These are cross-referenced by request format and by the software component that made the request. To see the round-trip times (RTTs) for all of these requests, you can use show statistics namespace ... response-detail. For a summary, showing request and response counts together with RTTs, use show statistics namespace ... summary. The output of all three commands includes the counts of various errors. To see the raw read/write statistics for the same namespace, volume, or share, you can use the show statistics namespace ... fastpath command. As with this command, you can see these statistics summarized for an entire namespace, focused on a single volume, or focused on one share.
Volume identify the volume.
All shares in volume appears if you did not use the share keyword to specify a single share. This indicates that the statistics are at a volume level; they are the sum of the statistics for all of the volumes shares. If you select a particular share, the following fields appear instead:
Share identifies the share, as it is named in the ARX configuration. The show global-config namespace name command shows the share names in the ARX configuration.
Filer is the name of the external filer that hosts the share. Again, this is the name of the filer in the ARX configuration; you can use the show external-filer filer-name command to see the IP address and/or SPN for filer-name.
CIFS Share is the name of the export or share at the filer. Both names appear if this is used as a multi-protocol share.
Reset... shows the last time these statistics were cleared. The software resets these statistics whenever someone reloads the chassis or uses clear statistics filer on the share.
Data Plane is fastpath traffic from the network software.
Ctrl Plane (short for Control Plane) is traffic from volume-group software.
Policy shows the requests from the policy engine, which performs migrations, shadow copies, and other policy-related functions.
Probe is the requests from the health-checking subsystem in the namespace software. This subsystem sends periodic requests to every back-end share that is connected to an ARX volume.
NFS RPC - Each row of the NFS table shows the total counts for a particular NFS-RPC request. Each column shows the source of the RPC request, as described above.
Errors appear in a separate table under the NFS statistics. The table contains the following counters, with the same columns as the RPC table:
FormatError is the number of malformed NFS packets received from the filer(s).
NetworkError counts the send and receive failures.
ResourceError is the number of times that the ARX software ran out of resources. Contact F5 Support if you see these errors.
RPCError counts any errors from the RPC layer.
TimeoutError is the number of timeouts waiting for an NFS response.
CIFS SMB - Each row in the CIFS table shows the total counts for a given CIFS (SMB) request. This has the same columns as the NFS tables above.
Errors appear in a similar table under the CIFS statistics, with the following fields:
FormatError is the number of malformed SMB (CIFS) packets received from the filer(s).
NetworkError counts the send and receive failures for SMB traffic.
ResourceError shows the number of times that the ARX software had insufficient resources to complete the transaction. Contact F5 Support if you see these errors.
SignatureError counts any SMB packets with missing or incorrect signatures. You can use the cifs filer-signatures command to change the SMB-signing policy for the current namespace.
SMBError is the number of SMB packets that failed with errors that are not directly related to any of the above errors.
TimeoutError is the number of timeouts waiting for an SMB response.
You can use the clear statistics filer command to clear the statistics from all variations of show statistics namespace. If you use that command to clear the statistics from one share in a volume but not another, the volume-level statistics in this command are inconsistent; one share may have statistics for 2 months and the other share may have statistics for only 5 minutes. To confirm that the volume has consistent statistics, run this command on each of the volumes shares and verify that the Reset... field has the same date and time for each of them.
stoweA> show statistics namespace lodges request-detail
bstnA> show statistics namespace wwmed volume /acct request-detail
bstnA> show statistics namespace wwmed volume /acct share all request-detail
stoweA> show statistics namespace lodges request-detail
bstnA> show statistics namespace wwmed volume /acct request-detail
bstnA> show statistics namespace wwmed volume /acct share all request-detail
Use the show statistics namespace ... response-detail command to show latency statistics between a namespace and its back-end shares.
show statistics namespace ns [volume vol [share {shr | all}]] response-detail
ns (1-30 characters) identifies the namespace.
vol (optional; 1-1024 characters) specifies one volume in the namespace.
shr | all is a required choice if you use the optional share keyword:
shr (1-64 characters) specifies one share in the volume.
all shows all shares in the volume. This creates a separate table for each share; if you omit the share all option, the output contains a single table showing the sums of all shares.
response-detail shows the round-trip times between various ARX software and back-end shares. This also shows error counts.
This command displays round-trip times (RTTs) between the chosen namespace, volume or share and its back-end filers. These are cross-referenced by request format and by the software component that made the request. To see the counts for all of these requests, you can use show statistics namespace ... request-detail. For a summary, showing request and response counts together with RTTs, use show statistics namespace ... summary. The output of all three commands includes the counts of various errors. To see the raw read/write statistics for the same namespace, volume, or share, you can use the show statistics namespace ... fastpath command. As with this command, you can see these statistics summarized for an entire namespace, focused on a single volume, or focused on one share.
Volume identify the volume.
All shares in volume appears if you did not use the share keyword to specify a single share. This indicates that the statistics are at a volume level; they are the sum of the RTT statistics for all of the volumes shares. If you select a particular share, the following fields appear instead:
Share identifies the share, as it is named in the ARX configuration. The show global-config namespace name command shows the share names in the ARX configuration.
Filer is the name of the external filer that hosts the share. Again, this is the name of the filer in the ARX configuration; you can use the show external-filer filer-name command to see the IP address and/or SPN for filer-name.
CIFS Share is the name of the export or share at the filer. Both names appear if this is used as a multi-protocol share.
Reset... shows the last time these statistics were cleared. The software resets these statistics whenever someone reloads the chassis or uses clear statistics filer on the share.
Data Plane is fastpath traffic from the network software.
Ctrl Plane (short for Control Plane) is traffic from the volume-group software.
Policy shows the RTT times from the policy engine, which performs migrations, shadow copies, and other policy-related functions.
Probe is the RTT times from the health-checking subsystem in the namespace software. This subsystem sends periodic requests to every back-end share that is connected to an ARX volume.
NFS RPC - Each row of the NFS table shows the average RTTs for a particular NFS-RPC request. Each column shows the source of the RPC, as described above.
Errors appear in a separate table under the NFS statistics. The table contains the following counters, with the same columns as the RPC table:
FormatError is the number of malformed NFS packets received from the filer(s).
NetworkError counts the send and receive failures.
ResourceError is the number of times that the ARX software ran out of resources. Contact F5 Support if you see these errors.
RPCError counts any errors from the RPC layer.
TimeoutError is the number of timeouts waiting for an NFS response.
CIFS SMB - Each row in the CIFS table shows the average RTTs for a given CIFS (SMB) request. The columns show the source of each SMB, as above.
Errors appear in a similar table under the CIFS statistics, with the following fields:
FormatError is the number of malformed SMB (CIFS) packets received from the filer(s).
NetworkError counts the send and receive failures for SMB traffic.
ResourceError shows the number of times that the ARX software had insufficient resources to complete the transaction. Contact F5 Support if you see these errors.
SignatureError counts any SMB packets with missing or incorrect signatures. You can use the cifs filer-signatures command to change the SMB-signing policy for the current namespace.
SMBError is the number of SMB packets that failed with errors that are not directly related to any of the above errors.
TimeoutError is the number of timeouts waiting for an SMB response.
You can use the clear statistics filer command to clear the statistics from all variations of show statistics namespace. If you use that command to clear the statistics from one share in a volume but not another, the volume-level statistics in this command are inconsistent; one share may have statistics for 2 months and the other share may have statistics for only 5 minutes. To confirm that the volume has consistent statistics, run this command on each of the volumes shares and verify that the Reset... field has the same date and time for each of them.
bstnA> show statistics namespace medarcv volume /rcrds response-detail
bstnA> show statistics namespace medarcv volume /rcrds response-detail
Use the show statistics namespace ... summary command to show request counts, response counts, and average round-trip times (RTTs) between a namespace and its back-end-filer shares.
show statistics namespace name [volume vol [share {shr | all}]] [summary]
name (1-30 characters) identifies the namespace.
vol (optional; 1-1024 characters) specifies one volume in the namespace.
shr | all is a required choice if you use the optional share keyword:
shr (1-64 characters) specifies one share in the volume.
all shows all shares in the volume. This creates a separate table for each share; if you omit the share all option, the output contains a single table showing the sums of all shares.
summary (optional) shows request and response counts per ARX request type, and shows the average RTT for each request type. This also shows error counts.
This command shows the request counts, response counts, and round-trip times (RTTs) from the chosen namespace, volume or share to its back-end filers. These are cross-referenced by request format and by the software component that made the request. To show the detailed request counts, you can use show statistics namespace ... request-detail. To see the detailed RTTs for all of these requests, you can use show statistics namespace ... response-detail. The output of all three commands includes the counts of various errors. To see the raw read/write statistics for the same namespace, volume, or share, you can use the show statistics namespace ... fastpath command. As with this command, you can see these statistics summarized for an entire namespace, focused on a single volume, or focused on one share.
Volume identify the volume.
All shares in volume appears if you did not use the share keyword to specify a single share. This indicates that the statistics are at a volume level; they are the sum of the statistics for all of the volumes shares. If you select a particular share, the following fields appear instead:
Share identifies the share, as it is named in the ARX configuration. The show global-config namespace name command shows the share names in the ARX configuration.
Filer is the name of the external filer that hosts the share. Again, this is the name of the filer in the ARX configuration; you can use the show external-filer filer-name command to see the IP address and/or SPN for filer-name.
CIFS Share is the name of the export or share at the filer. Both names appear if this is used as a multi-protocol share.
Reset... shows the last time these statistics were cleared. The software resets these statistics whenever someone reloads the chassis or uses clear statistics filer on the share.
NFS RPC - Each row of the NFS table shows the summaries for a particular NFS-RPC request. This has the following columns:
Req Count is the total number of requests from this namespace, including requests from the control plane, data plane, policy engine, and probe subsystem.
Resp Count is the total number of responses from this namespaces filers. These counts should match the request counts.
Avg RTT (uSec) is average round-trip time (RTT) for the request/response pairs.
Errors appear in a separate table under the NFS statistics. The table contains the following counters:
FormatError is the number of malformed NFS packets received from the filer(s).
NetworkError counts the send and receive failures.
ResourceError is the number of times that the ARX software ran out of resources. Contact F5 Support if you see these errors.
RPCError counts any errors from the RPC layer.
TimeoutError is the number of timeouts waiting for an NFS response.
CIFS SMB - Each row in the CIFS table shows the summaries for a given CIFS (SMB) request. This has the same columns as the NFS RPC table.
Errors appear in a similar table under the CIFS statistics, with the following fields:
FormatError is the number of malformed SMB (CIFS) packets received from the filer(s).
NetworkError counts the send and receive failures for SMB traffic.
ResourceError shows the number of times that the ARX software had insufficient resources to complete the transaction. Contact F5 Support if you see these errors.
SignatureError counts any SMB packets with missing or incorrect signatures. You can use the cifs filer-signatures command to change the SMB-signing policy for the current namespace.
SMBError is the number of SMB packets that failed with errors that are not directly related to any of the above errors.
TimeoutError is the number of timeouts waiting for an SMB response.
You can use the clear statistics filer command to clear the statistics from all variations of show statistics namespace. If you use that command to clear the statistics from one share in a volume but not another, the volume-level statistics in this command are inconsistent; one share may have statistics for 2 months and the other share may have statistics for only 5 minutes. To confirm that the volume has consistent statistics, run this command on each of the volumes shares and verify that the Reset... field has the same date and time for each of them.
stoweA> show statistics namespace lodges summary
stoweA> show statistics namespace lodges summary
This command applies only to a share in a multi-protocol (NFS and CIFS) volume with one or more NFS-only directories. An NFS-only directory has a name that cannot be expressed in both NFS and CIFS, so its back-end filer has created a filer-generated name (FGN) for CIFS clients. The CIFS attributes for NFS-only directories are obscured because it is not possible to deterministically find the directorys CIFS-side FGN. Directory duplication, called striping, is required for many standard operations in a managed volume, including file migration (place-rule, auto-migrate, balance) and share removal (remove-share migrate, remove-share nomigrate). The volume can never stripe an NFS-only directory if strict-attribute consistency is required. The no strict-attribute-consistency command allows striping of NFS-only directories; it permits the volume to ignore any CIFS attributes that it cannot find. The volume replaces any undiscovered CIFS attribute with a 0 (zero).
Use the affirmative form, strict-attribute-consistency, for multi-protocol volumes where NFS-only striping is not required.
Important: If you migrate to a share with no strict-attribute-consistency, NFS-only directories lose CIFS attributes at the destination share.
You may want to re-enable strict-attribute-consistency on the share after completing a one-time migration (such as a remove-share migrate).
The best practice in this case is to rename all NFS-only directories so that they are accessible to CIFS clients. If all directories are renamed, you can keep the strict-attribute-consistency setting for all shares in the volume and support all forms of migration and striping.
Additionally, NFS supports case collisions and CIFS does not. A case collision is two files or directories whose names differ only in case: for example, oneDir and ONEDIR.
Note that NFS clients can create NFS-only directories while the volume is in service. If strict-attribute-consistency is enabled, any such directory-creation (or rename) disables all striping for the directory. It also makes the directory and its contents inaccessible to CIFS clients, regardless of the strict-attribute-consistency setting.
bstnA(gbl-ns-vol-shr[insur~/claims~precs])# no strict-attribute-consistency
bstnA(gbl-ns-vol-shr[wwmed~/acct~bills])# strict-attribute-consistency
Use the wait-for remove command to wait until a volume-removal or share-removal operation has completed.
wait-for remove namespace vol-path [share] [timeout timeout]
namespace (1-30 characters) is the name of the namespace.
vol-path (1-1024 characters) identifies the volume.
share (optional, 1-64 characters) is the share that is being removed.
timeout (optional, 1-2096) is the timeout value in seconds.
timeout - none, wait indefinitely
When removing a volume with remove namespace ... volume, or a share with remove-share migrate or remove-share nomigrate, you can use the wait-for remove command to wait for the operation to complete. This can be useful for CLI scripts, which you can copy onto the switch (with copy ftp, copy scp, copy {nfs|cifs}, or copy tftp), and run.
If you set a timeout and it expires before the volume or share is removed, the command exits with a warning. To interrupt the wait-for remove command, press <Ctrl-C>.
bstnA> wait-for remove ns21 /vol4
bstnA> wait-for remove ns66 /vol31 rxbills timeout 60
remove namespace ... volume
Table of Contents   |   << Previous Chapter   |   Next Chapter >>

Was this resource helpful in solving your issue?




NOTE: Please do not provide personal information.



Incorrect answer. Please try again: Please enter the words to the right: Please enter the numbers you hear:

Additional Comments (optional)