Health Status And Recovery - IBM Storwize V7000 Unified Problem Determination Manual

Sample Output:
mgmt001st001 HOST_STATE
SERVICE
CTDB
GPFS
SCM
NETWORK
CHECKOUT
mgmt002st001 HOST_STATE
SERVICE
CTDB
GPFS
SCM
MGMTNODE_REPL_STATE OK
NETWORK
V7000 CLUSTER
ENCLOSURE
IO_GRP
MDISK
NODE
PORT
EFSSG1000I The command completed successfully.
If none of the previous steps resolved the GUI connectivity issues, perform the
following procedure.
Network port isolation for GUI:
If none of the previous steps have resolved the problem and the network
connectivity and system reports nothing wrong, there might be an issue with the
port configuration of your network that is not detected in any of the previous
steps. The internal management services use both port 443 and port 1081. Port 443
is redirected to port 1081 that the management service listens.
1. Check to see if you can access the GUI on the default https port (no port
2. Check network port settings and firewall settings. If the previous step fails,

Health status and recovery

Use this information to review the outstanding issues that cause the Health Status
indicator at the bottom of all management GUI panels to be red (critical errors) or
yellow (warnings or degradations).
– Yes: Run the CLI command lshealth. Reference the active
management node Hostname (mgmt001st001 or mgmt002st002)
obtained from the lsnode command. Ensure that HOST_STATE,
SERVICE, and NETWORK from lshealth is set to OK.
OK OK
OK All services are running OK
OK CTDBSTATE_STATE_ACTIVE
OK ACTIVE
OK SCM system running as expected
ERROR Network interfaces have a degraded state
OK Disk Subsystem have a online state
OK OK
OK All services are running OK
OK CTDBSTATE_STATE_ACTIVE
OK ACTIVE
OK SCM system running as expected
OK
ERROR Network interfaces have a degraded state
ERROR Alert found in component cluster
ERROR Alert found in component enclosure
OK The component io_grp is running OK
OK The component mdisk is running OK
OK The component node is running OK
ERROR Alert found in component port
– No: Perform network connectivity isolation procedures. Refer to
"Management node role failover procedures" on page 165.
No: Perform network connectivity isolation procedures. Refer to
"Management node role failover procedures" on page 165.
included in the URL). If all is good with firewall and management IP, the GUI
will listen on https://<Management IP>/ and provide a login prompt.
investigate the following issues:
v The firewall is open between the administrative browser and the Primary
Node Service IP but not between the administrative browser and the
management IP. The firewall settings must have rules that allow port 1081
but not 443 between the administrative browser and the management IP.
v The management IP is up but the port redirection on the switch/router is not
working as expected. Check the network settings.
Chapter 3. Getting started troubleshooting
23