Chapter 5 Cautions - Fujitsu ETERNUS User Manual

Chapter 5
This chapter discusses cautions that should be observed when using ETERNUSmgr.
Solaris OS, HP-UX, AIX, Linux, and Windows® Version Cautions
• Failure Notification
• When the administrator is informed by ETERNUSmgr of a failure in a target device, use the
ETERNUSmgr [Device Status Display] screen to find and confirm the status of the failed
component. After confirmation, contact your Fujitsu engineer and pass along this information.
• Whenever necessary, ETERNUSmgr outputs history HTML files and log files for each Alarm.
If the number of days to keep these files is set too high, or set to "0" (= "do not delete"), a
large number of files may build up over an extended period of use. Have the administrator
decide the appropriate number of days to keep these files.
• Network environment
• Polling and sending of E-mail may be impossible during a network failure on the
ETERNUSmgr server due to LAN drop-out and failure of communications equipment (e.g.
hubs, routers). During such failures, the target devices are not monitored, and if the Detect
Offline Targets ("DETECT_OFFLINE") flag is not set, the administrator may be unaware of the
failure. Enable the Detect Offline Targets parameter as necessary, and periodically check the
ETERNUSmgr log file (only for the log acquisition setting).
• ETERNUSmgr may not be able to respond to an event due to a target device or network
overload caused by a device failure. In such a case, refer to
Shutdown" (page 39)
the network environment is free from any problem.
• If the ETERNUSmgr server's network settings are not correct, or there are hardware failures
(LAN omissions, communications device failures, etc.) on the LAN path, it can take an overly
long time to execute polling or send E-mail, making periodic monitoring of target devices
impossible. In particular, in cases where a host name is specified for the target device to be
monitored rather than an IP address, and multiple DNS servers are registered for the
network, frequent inquiries must be made to the external network to resolve the host name.
This can take from several minutes to tens of minutes when the network is not working well.
• When it takes an overly long time to execute polling or send E-mail, check that the network
devices are operating normally, and that the network settings are correct.
• After rebooting a target device following a failure, etc., make sure that polling is successfully
being conducted by checking the log file. If polling is unsuccessful after the target device has
been rebooted (log information is not being output to the log file), refer to
and Shutdown" (page 39)
Even after ETERNUSmgr is rebooted, ETERNUSmgr polling requests may still not be
received by the target device (log information indicating polling failure is kept to be output to
the log file). In this case, the ETERNUSmgr monitoring function may not be working due to
network failure. Contact your Fujitsu engineer.
P2X0-0202-04ENZ0 ETERNUS Disk storage systems ETERNUSmgr User Guide -Introduction-
Cautions
and manually restart ETERNUSmgr. Before restarting, make sure that
and manually reboot ETERNUSmgr.
48
"Chapter 3 Startup and
"Chapter 3 Startup
Copyright 2009 FUJITSU LIMITED