| Title: | DEC Rdb against the World |
| Moderator: | HERON::GODFRIND |
| Created: | Fri Jun 12 1987 |
| Last Modified: | Thu Feb 23 1995 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 1348 |
| Total number of notes: | 5438 |
Hello, a customer of mine had an idea about 'fail-over'of ORACLE in a cluster and asked me if this could work. I didn't know the answer, so let me ask all of you. The situation: - 2 VAXes in a Cluster (one only for fail-over) - An ORACLE running on each node - Node 1 fails Imagine that there is a simulation of the DEADMAN-Lock of RDB via the Distributed Lock Manager, so that the ORACLE (or a watcher process) on the other node knows about the failure: Is it possible that the ORACLE on Node 2 does the recovery ? Any idea ? Thanks for input, Christian
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 665.1 | Nope | CLYPPR::BOOTH | What am I?...An Oracle? | Tue Jun 12 1990 15:42 | 10 |
Oracle V6 doesn't in any way use the Distributed Lock Manager. So this
scenario could be done if the customer were willing to write all his
own failover code.
Point two is that V6 may only have one active "instance" in a
VAXcluster for each Oracle RDBMS. That is, if database A is the one
being used, only one active Oracle RDBMS can function against that
database. So even a "live" failover database would fail to function.
---- Michael Booth
| |||||