Availability Manager Version 1.1 Release Notes The following notes address late-breaking information and known problems for the Availability Manager Version 1.1. o Recommended hardware configurations Compaq recommends using, at a minimum, either of the following hardware configurations: - 350 MHz Intel Pentium processor with 64 Mb of memory - 500 MHz OpenVMS Alpha processor with 128 Mb of memory o Some DECamds features not yet implemented The Availability Manager is, in most respects, a Java[R] implementation of the original DECamds availability management software product. However, some features of DECamds have not yet been implemented in the Availability Manager. These features are planned to be added in a future release. o Default data collection on OpenVMS nodes By default, the only data collected on OpenVMS nodes is node summary data. To view other data in an OpenVMS Node pane, you must select each type of data you want to collect on Data Collection pages. To use these pages, select the Customize VMS menu option in the Application window. o Limit your collection of detailed data By default, the only data collected on OpenVMS nodes is node summary data. You can collect this data on many nodes without incurring performance problems. However, restrict your collection of other data to a small number of nodes at a time. These performance problems will be addressed in a future release. o Avoid using multiple Data Analyzers on the same system Usually, a Data Analyzer displaying a few live nodes begins to show the nodes in the Application window within 10 to 20 seconds after the session starts. If you wait more than 2 minutes and still do not see a 1 group listed in the Group pane, most likely the driver has hung. The reason is probably that two sessions of the Availability Manager have overlapped, and the driver is not set up to handle multiple sessions. This problem is indicated if the Windows NT Task Manager shows a wjview.exe process when you think the session has ended. If this occurs, you need to reboot your system to clear the driver confusion. Also, do not run more than one Data Analyzer at a time. o Restarting the Availability Manager after an uninstall To uninstall the Availability Manager from a Windows NT system using Add/Remove Programs on the Windows NT Control panel, you must follow these steps: 1. Uninstall the software. 2. Restart the system. This step completes the removal of the network bindings. 3. Optionally, reinstall the software. If you omit Step 2 and perform Step 3, starting the Availability Manager could cause the system to fail. (This might occur even after the system appears to restart successfully.) To recover from this situation, reinstall the Availability Manager (uninstalling the software again is not necessary), and then restart your system at the end of the installation. The Availability Manager should run properly. o Large fonts cause incorrect displays If you specify large fonts in the Windows NT Display control panel, the text in some customize and fix dialog boxes does not display correctly. o Incorrect times displayed on the Cluster Summary page The times displayed on the Cluster Summary page differs by the number of hours your time zone differs from Greenwich Mean Time (GMT). (Depending on your location, this time difference could be added to or subtracted from your local time.) The reason is that the OpenVMS 2 time values are passed to the Data Analyzer as binary values and are not tagged with time zone indicators. Although a fix does not exist for this problem, a workaround is being investigated. o Blank rows in display trees Some displays of data in object trees might have blank rows because of irregularities in the underlying synchronization. This problem is more obvious in dual- processor or multiprocessor systems. Although a fix does not exist for this problem, a workaround is being investigated. Refer to the following website for updates to the current version of the software: www.openvms.digital.com/openvms/products/omt o Problems with event displays The following are known problems with the reporting of events: - Unimplemented threshold events LCKBLK NOPROC LCKWAT RESDNS LOSTVC RESPRS LRGHSH - Known threshold event problem Stopping the collection of node summary data does not clear the data from the Node pane. - Known event reporting irregularities * If a process goes away, posted events for the process might not be removed from the display. * LOVOTE events are posted for every node in the cluster rather than a single LOVOTE event for the entire cluster. Additional irregularities in the reporting of events might occur. o Windows NT Data Collector does not recognize new disk configurations 3 If you change the disk configuration on a running Windows NT node, the Data Collector does not recognize the modified disk configuration and continues to report the previous configuration to the Data Analyzer. For the Data Collector to recognize the new disk configuration, you must stop and restart the Data Collector. 4