Helpful isi CLI commands

isi_radish –q (helpful for finding bay-to-twe mapping and drive serial numbers)
fs61-7# isi_radish -q
Bay 1/twed0 is Maxtor 7Y250P0 FW:YAR41BW0 SN:Y62VM7AE, 490234752 blks
Bay 2/twed1 is Maxtor 7Y250P0 FW:YAR41BW0 SN:Y62VN2ME, 490234752 blks
Bay 3/twed2 is Maxtor 7Y250P0 FW:YAR41BW0 SN:Y62WDAKE, 490234752 blks
Bay 4/twed3 is Maxtor 7Y250P0 FW:YAR41BW0 SN:Y62VMYEE, 490234752 blks
Bay 5/twed4 is Maxtor 7Y250P0 FW:YAR41BW0 SN:Y62X1J4E, 490234752 blks
Bay 6/twed5 is Maxtor 7Y250P0 FW:YAR41BW0 SN:Y62Y1BXE, 490234752 blks
Bay 7/twed6 is Maxtor 7Y250P0 FW:YAR41BW0 SN:Y62VN57E, 490234752 blks
Bay 8/twed7 is Maxtor 7Y250P0 FW:YAR41BW0 SN:Y62VN6CE, 490234752 blks
Bay 9/twed8 is Maxtor 7Y250P0 FW:YAR41BW0 SN:Y62VN8ME, 490234752 blks

isi stat –n # (helpful for finding which bay the dead drive is in)
jerryb@TSUPPORT2-2:~$ isi stat -n 4
Node ID: 4
Node Name: TSUPPORT2-4
Node IP Address: 172.16.0.222
Node Health: [ATTN]
Node Capacity 931GB
Available 874GB ( 6%)
Used 57GB ( 6%)

Disk Drive Status:

twed4 <2> Empty Empty

0 b/s [DOWN] [DOWN]

twed5 <1> twed2 <4> twed0 <6>

0 b/s 0 b/s 0 b/s

twed6 <0> twed3 <3> Empty

0 b/s 0 b/s [DOWN]

isi_drivenum (Also shows bay-to-twe mapping)

Intravision-2# isi_drivenum
Bay 1 Unit 0 Lnum 8 Active /dev/twed0
Bay 2 Unit 1 Lnum 7 Active /dev/twed1
Bay 3 Unit 2 Lnum 6 Active /dev/twed2
Bay 4 Unit 3 Lnum 5 Active /dev/twed3
Bay 5 Unit 4 Lnum 4 Active /dev/twed4
Bay 6 Unit 5 Lnum 3 Active /dev/twed5
Bay 7 Unit 6 Lnum 2 Active /dev/twed6
Bay 8 Unit 7 Lnum 1 Active /dev/twed7
Bay 9 Unit 8 Lnum 0 Active /dev/twed8
 isi_net_info (helpful to ensure MTU is set correctly)
repfs01-2# isi_net_info
1500 bytes to 172.17.110.191 default mtu: 0% packet loss
1500 bytes to 172.17.110.191 real mtu: 0% packet loss
9000 bytes to 172.17.110.191 default mtu: 0% packet loss
9000 bytes to 172.17.110.191 real mtu: 0% packet loss
1500 bytes to 172.17.110.192 default mtu: 0% packet loss
1500 bytes to 172.17.110.192 real mtu: 0% packet loss
9000 bytes to 172.17.110.192 default mtu: 0% packet loss
9000 bytes to 172.17.110.192 real mtu: 0% packet loss
1500 bytes to 172.17.110.193 default mtu: 0% packet loss
1500 bytes to 172.17.110.193 real mtu: 0% packet loss
9000 bytes to 172.17.110.193 default mtu: 0% packet loss
9000 bytes to 172.17.110.193 real mtu: 0% packet loss

isi_hw_sane (NEW in 2.5.1. Shows ‘slow’ drives. Available on ftp.isilon.com as

/outgoing/bin/isi_hw_sane.)

Intravision-2# /root/isi_hw_sane

: total physical memory (3734417408 kB) looks okay

: isi_radish -F drive settings look okay
: drive speed (kB/s) 8kB read 8kB write 128kB read 128kB write
: bay 1 (A603SSZE) ok 39024:34000 35839:27000 39789:34000 40947:34000
: bay 2 (A603TA3E) ok 39797:34000 34749:27000 39730:34000 39723:34000
: bay 3 (A603SRPE) ok 39921:34000 34869:27000 40064:34000 40966:34000
: bay 4 (A603SWLE) ok 38932:34000 37609:27000 38944:34000 40125:34000
ERROR: bay 5 (Y60NK1FE) slow 40242:40000 46246:40000 54116:54000 62751:63000
: bay 6 (A603T5WE) ok 38732:34000 35812:27000 39934:34000 40619:34000
: bay 7 (A603SW9E) ok 38202:34000 35342:27000 39892:34000 40960:34000
: bay 8 (A603SW8E) ok 38978:34000 36890:27000 39559:34000 39915:34000
: bay 9 (A603SSBE) ok 39563:34000 36147:27000 39832:34000 38653:34000
----------
-- FAIL -- Saw 1 total hardware oddities
----------

isi_hwcheck (Show temps from the node. Note a “-1” means there was no reading from
the probe. /usr/bin/isi_hwtools/isi_hwcheck)

Intravision-2# /usr/bin/isi_hwtools/isi_hwcheck
Chassis: 9-Bay (AIC 9-Bay Chassis)
Mobo: Force4201 (Force 4201 Motherboard)
NVRam: MM5415 (MicroMemory 5415 33MHz NVRam Card) (size 536870912B)
DskCtl: TWE7500-12 (3Ware 7500 12-port Controller) (12 ports)
IMB: Board Version 3
Power Supplies OK
CPU Operation (raw 0x0) = Normal
Fan 1 (MB Fan 0) = -1
Fan 2 (MB Fan 1) = -1
Fan 3 (F1) = 4643
 Fan 4 (F2) = 4488
Fan 5 (F3) = 4691
Fan 6 (F4) = 4488
Voltage Vcc = 1.43
Voltage 3.3v = 3.30
Voltage 2.5v = 2.46
Voltage 5v = 4.91
Voltage 12v = 12.07
Voltage 5v (IMB) = 4.94
Voltage 12v (IMB) = 12.25
Temperature Management Chip = 30 Degrees C
Temperature Drives 1-6 (T1) = 23 Degrees C
Temperature Drives 7-9 (T2) = 22 Degrees C
Temperature PCI Riser Exhaust (T3) = 34 Degrees C
Temperature NIC area (T4) = 28 Degrees C
Temperature CPU Exhaust (T5) = 28 Degrees C
Temperature IMB = 25 Degrees C
Temperature CPU = 32 Degrees C

isi_for_array (Executes a command on every node of the cluster. ‘man isi_for_array’ is

really helpful. Pay particular attention to the –q, -l, and –s flags.)

root@TSUPPORT2-2:~$ isi_for_array -q uptime

Password:
TSUPPORT2-2: 10:13AM up 1 day, 21:28, 2 users, load averages: 0.02, 0.03, 0.01
TSUPPORT2-1: 10:13AM up 1 day, 21:28, 3 users, load averages: 0.21, 0.09, 0.02
TSUPPORT2-4: 10:13AM up 1 day, 21:28, 0 users, load averages: 0.01, 0.01, 0.00

isi_dmilog (Show the BIOS DMI log. Note, a POST error is normal. It means the
motherboard didn’t detect a keyboard on boot. Also, since the Bios log is full, no new
messages will be written. The log will have to be cleared before new messages will
be written.)

Intravision-2# isi_dmilog
Current change-id: 245
Note: Bios event log is full.
Log:
04/29/04 20:57:10 POST error: 00000200 00000000
04/29/04 21:36:09 POST error: 00000200 00000000
04/29/04 21:58:27 POST error: 00000200 00000000
04/29/04 22:03:03 POST error: 00000200 00000000
04/30/04 17:41:02 POST error: 00000200 00000000
04/30/04 19:11:29 POST error: 00000200 00000000
05/03/04 21:20:42 POST error: 00000200 00000000
05/03/04 21:34:45 POST error: 00000200
00000000
05/03/04 22:30:58 POST error: 00000200 00000000
05/03/04 22:50:46 POST error: 00000200 00000000
05/04/04 16:31:50 POST error: 00000200 00000000
05/04/04 16:52:56 POST error: 00000200 00000000
Totals:
12 of POST error

isi_bug_info (Way too much output to show a sample here. Engineering will sometimes
ask for this output. Here’s how to capture it.)

isi_bug_info > /tmp/demo-1.bug_info

isi_dav_cp – Copy utility that uses the dav protocol (http). This is very
useful when ftp is not good enough.

usage: dav_cp src1 [src2...] targethost:[targetport]/targetdir

-v: verbose - print out what is happeneing at the file level.
-d: debug - print out what is happening at the protocol level.
-u <user>: user - for authentication.
-p <pass>: password - for authentication.

isi_distill –The distill tool is used to capture the file and directory structure of a
customer’s data. This tool is useful to understand file size usage. It can also be useful to
determine empirically the performance of the cluster given a specific file structure.
Usage: isi_distill [-hq] [-o outfile] path
Options:
-h Print this message.
-p Preserve file names.
-q Be quiet about errors.
-o <outfile> Redirect output to <outfile> instead of stdout.

isi_dumpjournal – This tool is mainly used for debugging purposes. As

the name implies it dumps the content of the format into stdout.

Usage: isi_dumpjournal [-v] [-f <journal file>]

-v: Show transaction resolution archive as well
-f: Specify partition containing log (default /dev/mnv0)
-h: Print usage

isi_fixjournal –

isi_flush – Flushes the content of cache to disk (syncing to disk). This tool is useful when
trying to unmount the filesystem. Usually invoked by another script.

isi_get_itrace – Automated script for obtaining “itraces”. This script takes a 5 second
snapshot of performance information often required to debug problems with the cluster. It
generates a .gz file that is saved by default on /ifs/itraces/ directory.
isi_get_mbinfo – This script loads a debugging .ko to obtain mbuf information about a
node. It creates a .gz file that is saved by default on /ifs/itraces directory.

isi_restill – The restill tool is used to regenerate a file and directory structure using a distill
file. It can take a –j argument that parellizes the job of restilling files thereby making the
process faster.
Usage: isi_restill [-hiq] [-j job#/#jobs] basepath
Options:
-h Print this message.
-i <infile> Input file.
-j job#/#jobs Only do work for job# out of #jobs.
-n Ignore protection and coalescer settings.
-q Be quiet about errors.
-v Print files and directories as they are created.

isi_save_journal – This tools automates the process of saving the contents of the journal to
the /var/crash/ partition. This is one of the utilities used in the “disk tango” process.

isi_restore_journal – The restore_journal tool is used to recover a saved journal from the /
var/crash partition to the NVRAM.

isi_reimage – The isi_reimage tool installs a specified version of OneFS to a cluster or

node without preserving any of the cluster data. The reimage tool by default attempts to
mount buildbiox. Typing ‘ls’ at the prompt shows the images available for installation.
usage: isi_reimage [-mfh] [-b build]

-f: force to reimage without confirmatiom

-m: to reimage all the alive nodes on cluster
-i: ignore md5 checking after install
-h: help
-b build: to reimage with the build

isi_update - The isi_reimage tool installs a specified version of OneFS to a cluster or node
while preserving all of the cluster data. Unlike isi update, the tool by default attempts to
mount buildbiox. Typing ‘ls’ at the prompt shows the images available for upgrade
installation.

isi_kill_busy – The kill_busy tool attempts to kill any services that can cause prevent the
/ifs partition from being unmounted cleanly (filesystem is busy). This tool should only be
used when attempting to unmount the filesystem.

isi alerts
The isi alerts command provides access to active and historical alerts on the
cluster, and the ability to quiet, unquiet, or cancel any active alerts. To perform
one of the actions, specify the command followed by one or more alert ids or all.
>>> isi alerts [-w] [args [...]]

The isi alerts command with no argument appended returns a list of all currently
active alerts, with their ids, time of occurrence, severity, quiet status, and
message. If history is specified, then historical alerts are listed.

The -w argument specifies a wide view, to see all aspects of the alerts.

Alerts may be quieted if periodic notifications are no longer desired, or unquieted

to resume notifications. Alerts will be automatically cancelled after a period of time
if the condition no longer persists, or they may be cancelled explicitly by the user
(not recommended).

isi config
When a user executes isi config at the command line, the system opens a subset
of configuration commands directed at managing configuration settings. The isi
config command brings the user to an interactive configuration system where
various node and cluster settings may be altered.

NOTEWhile Isilon IQ is in the isi config subsystem, the other Isilon IQ commands
are unavailable.

>>> help

The help command lists all of the commands available in the current Isilon IQ CLI
context.

>>> help <Isilon IQ command>

The user can append the help command with another Isilon IQ command to get
details on how the specified command operates.

>>> close

The close command disconnects the user from the current isi_config session.

>>> config

The config command commits the latest configuration settings to the system.

>>> date [ < mm/dd/yyyy HH:mm:ss > ]

The date command sets the date and time the cluster uses for time/date
stamping.

If the user executes the date command with no arguments, it displays the current
system date and time. Or the user can set a new date and time using two-digit
 month (01-12) and day (01-31), and a four-digit year, and then a space to
separate the time using 24-hour format for hour values (00-23) and two-digit
values for minutes and seconds (00-60).

>>> dns [primary <IP> | secondary <IP> | domain <domain>] ...

The dns command, when executed with no argument, displays the DNS
information the cluster currently uses:

>>> dns
Primary DNS server is set to 172.16.0.1.

The secondary DNS server is not set.

The DNS domain is set to isilon.com.

>>> dns primary <IP>

Designates the IP address to use for the primary DNS server.

>>> dns secondary <IP>

Designates the IP address to use for the secondary DNS server.

>>> dns domain <IP>

Designates the domain name to use.

>>> dns delete [primary | secondary | domain]

The dns delete command deletes all DNS information if used without any other
argument. If delete is followed by primary, secondary or domain, only that
information is deleted.

>>> exit

The exit command disconnects the user from the current isi_config session.

>>> iprange [<interface-name> [<lowest IP>-<highest IP>]]

The iprange command displays a list of ranges of IP addresses currently

assignable to
nodes, when executed with no argument. It designates IP addresses assignable to
nodes, when appended with a range of IP addresses.
>>> list [clusters | unconfigured]

The list command displays a list of clusters on the same subnet as the node
executing the command, when appended with the 'clusters' argument.

It displays a list of unconfigured nodes on the same subnet as the node executing
the command, when appended with the 'unconfigured' argument.

When executed with no argument, it displays both lists.

The add and join commands use the list command.

>>> add < MAC address | list position >

The add command attempts to export the configuration settings from the cluster
executing the command to an unconfigured node on the same subnet, thus making
it part of the cluster. The node is specified in the argument by its MAC address, or
by the number corresponding to its position in the list of unconfigured nodes
(obtained by executing the 'list unconfigured' command). Multiple machines can be
added simultaneously.

>>> join < name | list position > (<password>)

The join command attempts to join a node to a cluster, specified by its name or by
the number that corresponds to its position in the list of clusters (see list).

>>> joinmode [ mode ]

The joinmode command determines how the system will behave when new Isilon
IQ nodes come online on the subnet occupied by the Isilon IQ cluster:

The manual argument configures the system to join new nodes into the cluster in
a separate manual process (see join and add). This is the default join mode
setting.

The secure argument configures the cluster to disallow any new node from joining
the cluster externally. It also makes some other aspects of operation more secure.

>>> migrate [ <interface-name> [<old IP range>] <new IP

range>]

The migrate command displays a list of ranges of IP addresses currently

assignable to nodes, when executed with no argument. It migrates all nodes that
are assigned addresses in the old range of IP addresses over to addresses in the
new range of IP addresses. If the old range is not given, all nodes in the cluster are
migrated to addresses in the new range.

>>> remove [<node id> | all]

The remove command deletes a node from the cluster, as specified by its node ID.
 If no argument specifying a node is appended, remove will act upon the node that
is running the current isi_config session.

If the 'all' argument is appended, it removes all cluster nodes. In all cases, the
system returns a confirmation prompt before executing the command.

Once executed, any removed node will reboot and return to its unconfigured state
with all data removed.

The system returns a warning prompt, asking the user to confirm execution:

DANGER: Resetting device 2 might cause data loss if there are

files on that
device which are not on any other device.
Are you sure you wish to continue? [no]
>>> < yes | no >

The default entry at this prompt is [no]; so if the user presses Enter without
typing a response, remove will not execute.

>>> deliprange [<interface-name> lowest IP - highest IP ]

The deliprange command removes the IP addresses in the range specified in the
argument from the list of IP addresses reserved for use in the particular named
interface.

If deliprange is executed with no arguments, the system will return a list of IP

ranges reserved for assignment to Isilon IQ nodes

If an address in the range specified to be deleted is currently assigned to an Isilon

IQ node, the system will return an error message:

>>> deliprange Primary 172.16.20.91-172.16.20.99

IP '172.16.20.91' is in use and may not be deleted

When the system has successfully executed the deliprange command, it returns a
confirmation message:

Removed IP range Primary 172.16.20.95-172.16.20.99.

>>> interface [<interface-name> [enable|disable]]

The interface command displays the IP ranges, netmask and MTU of all network
interfaces when issued with no argument. It displays the IP ranges, netmask and
MTU of the given interface when issued with a network interface name argument.
It enables or disables the given network interface when issued with the enable or
disable argument.
>>> name < name >

The name command displays the name currently assigned to the Isilon IQ cluster,
when executed with no argument. It sets the name of the cluster, when appended
with a name. The name will appear in the list of masters and should be used if the
cluster is named on a DNS server.

>>> netmask [<interface-name> [<x.x.x.x>]]

The netmask command displays the subnet IP mask that the named interface is
currently using, when executed with no argument. It sets the subnet IP mask that
the named interface will use, when appended by a valid subnet IP mask.

>>> gateway [<IP address> | delete]

The gateway command displays the IP address of the gateway machine the
cluster users to communicate with systems external to its Ethernet subnet, when
executed with no argument. Designates the gateway machine the cluster uses to
communicate with systems external to its Ethernet subnet, when appended with
the IP address of the gateway machine. Removes the gateway machine the cluster
currently uses, when appended with the 'delete' argument.

>>> mtu [<interface-name> [<value>|mixed]]

The mtu command displays the size of the Maximum Transmission Unit (MTU) the
named interface uses in network communications, when executed with no
argument. Sets the size of the MTU the named interface uses, when appended with
a value. Common settings are 1500 (standard frames), 9000 (jumbo frames), and
Mixed (1500 external, 9000 internal).

It is highly recommended that users configure the network interface used for
internal communications for Jumbo Frames. Jumbo Frames enable the Isilon IQ
cluster to more efficiently communicate to all storage nodes within the cluster,
improving read and write performance up to 30 percent.

NOTEChanges to the MTU setting will not be applied until the nodes are rebooted.

>>> quit

The quit command disconnects the user from the current isi_config session.

>>> reboot [<node id> | all]

The reboot command reboots the node from which the command is executed,
when executed with no argument. Reboots a specified node, when appended with a
node's ID. Reboots all cluster nodes, when appended with the 'all' argument.
>>> rediscover

The rediscover command attempts to reacquire Isilon IQ nodes on the network by

rebroadcasting discover packets.

>>> status [ advanced ]

The status command returns the current configuration of the cluster.

With the advanced argument appended, status gives additional configuration

details, such as the Master Node ID and the Local ifsd PID.

>>> timezone [<timezone identifier>]

The timezone command sets the timezone on the cluster to the specified
timezone. Valid timezone identifiers are: GMT, Greenwich Mean Time, Eastern,
Eastern Time Zone, Central, Central Time Zone, Mountain, Mountain Time Zone,
Pacific, Pacific Time Zone, Arizona, Alaska, Hawaii.

>>> virtip [add|delete [<ip address>]]

The virtip command displays the configured virtual IP address when issued with no
argument. Adds a new IP address to the list of configured virtual IP addresses
when appended with 'add' and an IP address. Removes an existing IP address from
the list of configured virtual IP addresses when appended with 'delete' and an IP
address.

>>> version

The version command displays product version information.

>>> wizard

The wizard command puts the user into a scripted series of commands that
manipulate the basic configuration settings.

isi get
The isi get command displays the the properties of the files and directories stored
in the Isilon IQ cluster, including their policy, protection level, and caching level. It
also contains color coded status information.

>>> isi get [-adDR] file

Isilon-1# isi get //ifs/test
POLICY LEVEL COAL FILE
+1 1+1 on 1g
+1 2x on Copy of dash/
2x 2x on dash/
 -aWhen appended with the -a flag, the isi get command is used to include the
hidden `.' and `..' entries of each directory and so displays the directory properties
as well as the file properties.
-dWhen appended with the -d flag, the isi get command is used to view the
attributes of a directory itself rather than its contents and displays directories as
plain files.
-DWhen appended with the -D flag, the isi get command is used for more
verbose information for debugging purposes and displays the data block IADDRS
where the data is located on the disk.
-RWhen appended with the -R flag, the isi get command is used to descend into
a directory recursively and displays the properties of the directory specified in the
path and any files and directories it contains.

At any level of output, the protection policy appears in one of three colors: green,
yellow, or red. The color indicates the protection status of the file, with green
corresponding full protection, yellow corresponding to a degraded protection under
a mirroring policy, and red corresponding to a loss of one or more data blocks
under a parity policy.

More succinctly, yellow indicates that a file "needs reprotect", while red indicates
that it "needs repair".

isi services
The isi services command controls the enabling or disabling of the various kinds of
services that are available in the Isilon environment, including NTP, smbtime, NIS,
NFS, Telnet, SSh, HTTP, Samba, FTP, and two Isilon utilities, isi spy and isi
restripe.

>>> isi services [-la] [service [<enable|disable>]]

The isi services command with no argument appended returns a list of the
activation status of each service:

Available Services:

ntpd Network Time Server Enabled

smbtime Windows Time Service Enabled
nfs NFS Server Enabled
nis Network Information Service Enabled
telnetd Telnet Server Enabled
sshd Secure Shell Server Enabled
httpd Web Server Enabled
samba Windows Share (SMB) Server Enabled
isi_spy_d Isilon Spy Daemon Enabled
ftp FTP server Disabled
isi_restripe_d Isilon Restripe Daemon Enabled
 If the isi services command is used with the enable argument, the service will be
turned on.

>>> isi services ftp enable

If the isi services command is used with the disable argument, the service will be
turned off.

>>> isi services ftp disable

-lWhen appended with the -l flag, the isi serv command will display all services
that the user may change, along with a short description and their state.

-aWhen appended with the -a flag, the services list will be augmented to include all
services and their current status, even those that the user cannot change.

isi set
The isi set command controls the protection policy settings on files and folders in
the Isilon IQ file system. There are several ways a protection policy can be set on a
file:

• a policy can be set explicitly on files

• a policy can be inherited from a parent folder (i)

• a policy on a folder can be forced upon the contents of that folder,

overriding the settings already in place.

>>>isi set [-fnrvR] [-p policy] [-c <on|off>] file

Protection Policy
To set a protection policy explicitly on a particular file, execute the isi set
command appended with the –p flag followed directly with a protection policy, and
then the path to the file on which you want to set the policy, as in the following
example:

<cluster name># isi set -p< policy > <file path>

Or, more specifically:

<cluster name># isi set –p[ i | <n>x | <n>+1 ] <file path>

Where the policy options act in the following ways:

• The i argument configures the file to inherit its protection policy from the
nearest parent folder that has an explicitly set policy. The i setting is the
implicit state.

• The <n>x argument configures the file system to mirror the file n times,
where n cannot exceed the number of nodes in the cluster.

• The <n>+1 argument configures the file system to stripe the file n
times with parity protection.

For example, the following command would protect the designated file with twice-
over mirroring:

<cluster name># isi set –p2x //ifs/song.mp3

Write-Coalescing Policy
Write-coalescing may be set to either on or off , and is specified with the -c flag.
By default, only the files specified on the command-line are modified, as in the
following example:

<cluster name># isi set -c on <file path>

Scope
The file upon which the protection policy acts can be further defined by prepending
the file path with one of the following scope operators/delimiters:

<cluster name># isi set [-fnrvR] –p<policy> <file path>

Where the scope options act in the following ways:

• The -f argument may be specified to suppress warnings on failures to

change a file.

• The -n argument outputs the list of files that would be changed without
actually taking any action.

• The -r argument explicitly forces a restripe. It relaxes the requirement

that a change be performed, and restripes each file regardless of whether
this would have occurred otherwise.

• The -v argument causes set to list each file as it is reached.

• The -R argument recursively sets the protection on files.

Summary
In its entirety, the isi set command takes this form:

isi set [-fnrvR] [-p policy] [-c <on|off>] file

isi status [-qwD] [-n [ID]]

When a user executes isi status at the command line of any node in an Isilon IQ
cluster, the user can get a snapshot of the system's current state.

The isi status command returns real-time information about the status of the
cluster, including data on the health, capacity, space available, detailed throughput
data, and operations whether finished, active, waiting, or failed. The Cluster Health
and Node Health status indicators are color coded, with green indicating healthy,
red indicating offline, and yellow indicating attention needed.

Cluster Name: Isilon

Cluster Health: [ OK ]
Cluster Capacity: 4.2TB
Available: 4.2TB (100%)
Used: 6.3GB ( 0%)
Throughput (bits/s)
ID | IP Address | Health | In Out Total | Used /
Total
----+-----------------+--------+-------+-------+-------
+------------------------
1 | 172.16.0.131 | [ OK ] | 0 | 0 | 0 | 2.1GB /
1.4TB ( 0%)
2 | 172.16.0.132 | [ OK ] | 0 | 0 | 0 | 2.1GB /
1.4TB ( 0%)
3 | 172.16.0.133 | [ OK ] | 0 | 0 | 0 | 2.1GB /
1.4TB ( 0%)

Finished Operations (2):

OPERATION START COMPLETE ELAPSED ERRORS

AutoBalance 12/05 17:21 12/05 17:21 00:00:05 0
Collect 12/05 17:21 12/05 17:21 00:00:03 0

No active operations.
 No waiting operations.

No failed operations.

The isi status -q flag ("quiet" output) is used to return the cluster status without
information on operations.

The isi status -w flag ("wide" output) specifies that no truncation of operations
status should take place.

The isi status -n flag ("node" output) returns information about each node in the
cluster. By using this command a user can quickly tell which nodes and/or disks
are up and down, as in the following example:

isi stat -n
Node ID: 1
Node Name: Isilon-1
Node IP Address: 172.16.0.131
Node SN:
Node Health: [ OK ]
Node Capacity 1.4TB
Available 1.4TB ( 0%)
Used 469MB ( 0%)

Network Status:

Prim / bc0
172.16.0.131
26K / 5.4K

Disk Drive Status:

twed5 <2> twed2 <5> ad0 <8>

0 b/s 13Kb/s 13Kb/s

twed6 <1> twed3 <4> twed0 <7>

0 b/s 0 b/s 13Kb/s

twed7 <0> twed4 <3> twed1 <6>

13Kb/s 0 b/s 0 b/s

Node ID: 2
Node Name: Isilon-2
Node IP Address: 172.16.0.132
Node SN:
Node Health: [ OK ]
Node Capacity 1.4TB
Available 1.4TB ( 0%)
Used 468MB ( 0%)

Network Status:

Prim / bc0
172.16.0.132
27K / 5.1K

Disk Drive Status:

twed5 <2> twed2 <5> ad0 <8>

1.6Kb/s 41Kb/s 1.6Kb/s

twed6 <1> twed3 <4> twed0 <7>

41Kb/s 41Kb/s 1.6Kb/s

twed7 <0> twed4 <3> twed1 <6>

1.6Kb/s 41Kb/s 1.6Kb/s

Node ID: 3
Node Name: Isilon-3
Node IP Address: 172.16.0.133
Node SN:
Node Health: [ATTN]
Node Capacity 1.2TB
Available 1.2TB ( 0%)
Used 454MB ( 0%)

Network Status:

Prim / bc0
172.16.0.133
34K / 4.7K

Disk Drive Status:

twed5 <1> twed2 <4> ad0 <7>

0 b/s 0 b/s 13Kb/s

Empty twed3 <3> twed0 <6>

[DOWN] 0 b/s 0 b/s

twed6 <0> twed4 <2> twed1 <5>

13Kb/s 13Kb/s 13Kb/s
Procedure - Softfail

1) Soft Fail the drive. This is what would occur in 99% of the cases
of a drive falling out of the acceptable Isilon specifications and
needing to be removed before a hard failure of the drive.

a. From the command line type the following:

i. >> isi conf
b. From within the isi configuration menu display the help
mode for softfail.
i. >>> softfail help
Unknown option: help

usage: softfail -q <node-drive-spec> : Query current

soft-fail status.
softfail -s <node-drive-spec> : Set soft-fail flag.

<node-drive-spec> ::= <node> [ ":" [ <drive> ] ]

::= [ <node> ] ":" <drive>

<node> ::= <number>

<drive> ::= "lnum" <number>

::= "bay" <number>
::= "port" <number>
::= "twed" <number>
::= "unit" <number>

<number> ::= [0-9]+

c. Use the help menu as your syntax guide select the
appropriate node, and drive you wish to softfail.

Some notes about this procedure:

1) This will put the selected drive into softfail mode as would the
OS or Isilon Customer Service determined the drive was no
longer acceptable to be part of the overall cluster. The system
will begin a restripe of the contents on this drive and re-stripe
the contents to other drives in the cluster. Once the restripe is
complete the drive is then logically removed the system.
 Please note: during this restripe the drive can still be read
from, but is no longer available for writes.
2) Repeat the procedure for another drive if you have N+2 set for
the entire system as you have planned.