This section describes commands that manage the set of servers within a grid.
Disable a server for scheduled maintenance.
When a server is in maintenance mode, the grid does not use the server at all. A grid maintainer typically puts a server in maintenance mode when they expect a server to be down for a long time or rebooted multiple times, such as for hardware changes or BIOS update, and they don’t want the grid to do anything with that server.
server disable ( <name> | --all ) [--maintenance] [reason=<reason>]
Name of server to disable
Disable all servers
Disable the server and place it into maintenance mode
Reason why server is being disabled
server disable srv2
Disable server 'srv2'
server disable --all
Disable all servers
server disable srv2 --Maintenance reason="Hardware upgrade"
Disable server 'srv2' for hardware upgrade
Notes:
Enable a server for scheduling.
server enable ( <name> | --all )
Name of server to enable
Enable all servers
server enable srv2
Enable server 'srv2'
server enable --all
Enable all servers
Notes:
Identify server and NICs by having the specified NIC on the server blink/flash its LED for a minute.
server identify name nic=nic-name
name of server
name of NIC whose LED is to blink/flash
server identify srv1 nic=eth1
Make LED for eth1 on srv1 blink for 1 minute
server identify srv2 nic=eth3
Make LED for eth3 on srv2 blink for 1 minute
Retrieve detailed information for a specific server.
server info name [ --batch ] [--extended] [ --map ]
Name of server for which to retrieve information.
Display information in UDL format
Display extended server information
Display list of components running on the server
server info srv1
Display information for 'srv1'
Notes:
Name : srv1 State : up HA Role : primary --- CPU Information --- Type : Intel(R) Xeon(R) CPU E5320 @ 1.86GHz Frequency : 1861 Hz Bogomips : 3725 Load : 0.49 --- Resource Information --- Resource Count Total Reserved Alloc Service Free ------------------------------------------------------------------------ CPU 2 8 0.10 0.14 - 7.76 MEM - 32.00 GB 1.44 GB 448.00 MB - 30.12 GB BW - 2.00 Gbps 0.00 Mbps 14.00 Mbps - 1.99 Gbps DISK 1 1.59 TB 10.00 GB 82.62 GB - 1.50 TB
Server Info Output
The output of the command output specification is as follows:
Name : val State : val | val (disabled) if server is disabled Down Reason : val (only if state is not up and not empty) Down User Responsible : val (only if state is not up and not empty) Disabled Reason val (only if server is disabled, not empty, and if --extended) Disabled User Responsible : val (only if server is disabled, not empty, and if --extended) Reboot Required : val (only if flagged for reboot) HA Role : val UID : val (only if --extended) Current Time : val (only if --extended) Uptime : val (only if --extended) Boot ID : val (only if --extended) Management Enabled : val (only if --extended) Bandwidth Oversubscription Enabled : val (yes/no - only if --extended) Disk NCQ Enabled : val (yes/no - only if --extended) Jumbo Frames Mode : val (enabled/disabled/auto - only if --extended) Jumbo Frames Enabled : val (yes/no – only if --extended)
SNMP Enabled : val (yes/no – only if --extended)
--- IP Information --- (only if maintainer)
Private : val
Public : val
--- Network Information --- (only if maintainer and --extended)
Network HA Mode NIC Switch State MAC
----------------------------------------------------------------------------
backbone none| ethN <switch-id> up|down|active xx:xx:xx:xx:xx:xx
active-passive| ethN <switch-id> up|down|active xx:xx:xx:xx:xx:xx
active-active
…
external none| ethN <switch-id> up|down|active xx:xx:xx:xx:xx:xx
active-passive| ethN <switch-id> up|down|active xx:xx:xx:xx:xx:xx
active-active …
-- Switch Information --- (only if maintainer and --extended)
ID Name Detection Method
---------------------------------------------------
<id> <name> <val>
…
-- BIOS Information --- (only if --extended)
Vendor : val
Version : val
Date : val
-- Motherboard Information --- (only if --extended)
Manufacturer : val
Model : val
Version : val
--- CPU Information ---
Type : val
Frequency : val
Bogomips : val
Load : val
Hyperthreading : on|off
--- Disk Check Information ---
Supported : val (displayed only if --extended is specified)
Enabled : val (displayed only if --extended is specified)
--- Resource Information ---
Resource Count Total Reserved Alloc Hypervisor Free
--------------------------------------------------------------------------------
CPU 1 <val> <val> <val> - <val>
MEM - <val> MB <val> MB <val> MB <val> MB <val> MB
BW - <val> Gbps <val> Gbps <val> Gbps - <val> Gbps
DISK 1 <val> GB <val> GB <val> GB - <val> GB
If --map is specified, a list of application components running on the server displays with the following information specified for each component:
--- Components ---
(Table with the following headings: Name, State, CPU, Memory(MB), Bandwidth(Mbps)
If -- batch is specified, the output is as follows:
server <name>
{
state = val
down_reason = val (only if state is not up and not empty)
down_user_responsible = val (only if state is not up and not empty)
enabled = val
disabled_reason = val (only if server is disabled, not empty, and if --extended)
disabled_user_responsible = val (only if server is disabled, not empty, and if --extended)
reboot_required = val
ha_role = val
uid = val (only if --extended)
current_time = val (only if --extended)
time_synced = val (only if --extended)
uptime = val (only if --extended)
boot_id = val (only if --extended)
mgmt_enabled = val (only if --extended)
bw_oversubscription_enabled = val (only if --extended)
disk_ncq_enabled = val (only if --extended)
jumbo_frames_mode = val (only if --extended
jumbo_frames_enabled = val (only if --extended)
snmp_enabled = val (only if --extended)
ip1 = xxx.xxx.xxx.xxx (only if maintainer)
ip2 = xxx.xxx.xxx.xxx (only if maintainer)
supported_virt_modes = val
network backbone (only if maintainer and --extended)
{
ha_mode = val
nic ethN : switch = <id>, mac = val, state = val
nic ethN : switch = <id>, mac = val, state = val
…
}
network external (only if maintainer and --extended)
{
ha_mode = val
nic ethN : switch = <id>, mac = val, state = val
nic ethN : switch = <id>, mac = val, state = val
…
}
switches
[
<id>=<val>, name=<val>, manufacturer=<val>, model=<val>, detection_method=<val>
…
]
bios_vendor = "val" (only if --extended)
bios_version = "val" (only if --extended)
bios_date = "val" (only if --extended)
motherboard_manufacturer = "val" (only if --extended)
motherboard_model = "val" (only if --extended)
motherboard_version = "val" (only if --extended)
hvm_support = val
cpu_phy_total = val
cpu_total = val
cpu_type = "val"
cpu_freq = val
cpu_bogomips = val
cpu_load = val
cpu_ht_enabled = val
cpu_reserved = val
cpu_free = val
cpu_alloc = val
mem_total = val
mem_reserved = val
mem_free = val
mem_alloc = val
mem_service = val
bw_total = val
bw_reserved = val
bw_free = val
bw_alloc = val
bw_oversubscription_enabled = val (only if --extended)
n_disks = val
disk_total = val
disk_reserved = val
disk_free = val
disk_check_supported = val (only displayed if --extended is specified)
disk_check_enabled = val (only displayed if --extended is specified)
If --map is specified, the following is displayed for each component/system VM running on the server
component <name>: state = val, cpu = val , mem = val , bw = val
}
List servers
Display output in UDL format
Display all server information in the list
Include extended server information - only valid if --verbose is specified
Display list of components running on each server
server list
List all servers of the grid.
Note:
Name State CPU Mem(MB) BW (Mbps) Role
Alloc Free Alloc Free Alloc Free
srv1 up 0.14 7.76 448 30841 14 1986 primary
srv2 up 0.00 8.00 0 31802 0 2000 secondary
This topic contains command output specification.
Non-batch output
Table with the following headings: Name State CPU Alloc/Free Mem(MB) Alloc/Free BW(Mbps) Alloc/Free Role Non-batch output (with --map) server name : role val, state val(enabled/disabled), val cpu, val MB mem, val Mbps bw srv list -m server srv1 : role primary, state up(enabled), 0.05/7.85 cpu, 128/30983 MB mem, 100/1900 Mbps bw _sys.controller: running, 0.10 cpu, 768.00 MB, 0.00 Mbps app1.main.LUX5: running, 0.05 cpu, 128.00 MB, 100.00 Mbps server srv2 : role secondary, state up(enabled), 0.00/8.00 cpu, 0/31880 MB mem, 0/2000 Mbps bw
server name: role = val, state = val, enabled = val, cpu_alloc = val,
cpu_free = val, mem_alloc = val, mem_free = val,
bw_alloc = val, bw_free = val, reboot_required = val
{ (if -map specified)
component name : state = val, cpu = val , mem = val, bw = val
…
}
and a list of application components running on the server:
name : val CPU, val MB, val Mbps
If --batch is specified on the command line, the output of this command is the following:
server name : role=val, state= val , enabled=val, cpu_alloc=val, cpu_free=val, mem_alloc=val, mem_free=val, bw_alloc=val
bw_free=val
{
component name : state=val, cpu=val, mem=val, mem_hypervisor=val, bw=val
}
If --verbose is specified on the command line, the output of this command is the same as the server info command except that the information is displayed for each server.
Initiate network topology discovery for a server.
server net_discover name | --all
Name of server whose network topology is to be discovered.
Initiate network topology discovery for all servers.
server net_discover srv1
Initiate network topology discovery for 'srv1'.
server net_discover --all
Initiate network topology discovery for all servers.
Notes:
Power cycle a server.
server power_cycle name [ reason=reason ] [ --kill ] [ --force ]
Name of the server to power cycle. If name is "--all", then all servers are power cycled.
Reason why server is being power cycled
Force power cycle of server even though there may be an application currently running on the server. This option is supported in CA AppLogic 2.8.x+.
Skip user verification. Prior to CA AppLogic 2.8.x, this option had the same functionality as the --kill option.
server power_cycle srv3
Power cycle 'srv3'
Note: This command will fail if there is an application currently running on the server (a volume is in use or a component is running on the server) and if the --kill option is not specified.
Power off a server.
server power_off name [ reason=reason ] [ --kill ] [ --force ]
Name of the server to power off. If name is "--all", then all servers are powered off.
Reason why server is being powered off
Force power off of server even though there may be an application currently running on the server. This option is supported in CA AppLogic 2.8.x+.
Skip user verification. Prior to CA AppLogic 2.8.x, this option had the same functionality as the --kill option.
server power_off srv3 reason="Memory upgrade"
Power off 'srv3'
Note: This command will fail if there is an application currently running on the server (a volume is in use or a component is running on the server) and if the --kill option is not specified.
Power on a server.
server power_on <name> | --all
Name of the server to power on.
--all - Power on all servers that are currently powered down.
server power_on srv2
power on 'srv2'
server power_on --all
power on all servers that are powered down
Reboot a server.
server reboot <name> | --all [ reason=<reason> ] [ --force ] [ --kill ] server reboot <name> | --all --flag
Name of server to reboot
Reboot all servers
Reason why server is being rebooted
Force reboot of server even though there may be an application currently running on the server.
Skip prompting user for verification
Flag the server for reboot (do not actually reboot)
server reboot srv2 reason="stuck mount"
reboot 'srv2'
server reboot --all
reboot all servers
server reboot srv2 --flag
flag srv2 for reboot
Set server configuration parameter.
server set <name> <prop>=<val> [ --force ]
Name of server for which to set information.
Server parameter to modify. The following parameters may be set:
Server's HA role. May be one of the following:
Server that is currently running the grid controller
Server that may run the grid controller in case of a controller server failure
Server will never run the grid controller and does not participate in controller HA
Enable/disable network HA for the server on both external and backbone
Enable/disable network HA on the backbone network
Enable/disable network HA on the external network
specify which switch is active for the backbone network
Specify which switch is active for the external network
Enable/disable NCQ
Enable or disable bandwidth oversubscription
Specify the volume repair speed. May be specified with ‘K’, ‘KB’, ‘M’, or ‘MB’ modifier to represent KB/sec or MB/sec . The valid range is between 1M and 200M inclusive. If no units modifier is specified, KB/sec is assumed.
Specify the number of CPU cores to set for scheduling. May be specified as an explicit number or a multiplier of the physical available resources, such as 1x or 2x. If empty, the number of available CPU cores is the same as the available physical CPU resources.
Skip the user verification prompt when setting a new primary server for a grid
server set srv1 role=primary
Set server srv1's role to primary
server set srv2 role=secondary
Set server srv2's role to secondary
server set srv1 ha_network=1
Enable network HA for the server (external and backbone)
server set srv2 ha_backbone=0
Disable network HA on the backbone network for the server
server set srv2 ha_external=0
Disable network HA on the external network for the server
server set srv2 backbone="11:22:33:44:55:66"
Set backbone NIC connected to switch with an ID of 11:22:33:44:55:66 as active, all backbone network traffic on srv2 passes through the NIC connected to that switch
server set srv2 external="77:88:99:AA:BB:CC"
Set external NIC connected to switch with an ID of 77:88:99:AA:BB:CC as active, all external network traffic on srv2 passes through the NIC connected to that switch
grid set overcommit.cpu_sched_limit=2x
Set the number of CPU cores for scheduling to twice the physical number of CPU cores.
Notes:
Shutdown a server.
server shutdown name [ reason=reason ] [ --kill ] [ --force ]
Name of the server to shutdown. If name is "--all", then all servers are shutdown.
Reason why server is being shutdown
Force shutdown of server even though there may be an application currently running on the server. This option is supported in CA AppLogic 2.8.x+.
Skip user verification. Prior to CA AppLogic 2.8.x, this option had the same functionality as the --kill option.
server shutdown srv3 reason="Memory upgrade"
Shutdown 'srv3'
Note: This command will fail if there is an application currently running on the server (a volume is in use, or a component on the server) and if the --kill option is not specified.
|
Copyright © 2013 CA Technologies.
All rights reserved.
|
|