Informix - Problem description
Problem IT34246 | Status: Closed |
LOG RECOVERY ERRORS ON SECONDARIES IF RESTARTED TOO SOON AFTER INSTATIATION FROM PHYSICALLY RESTORED NON-WHOLE-SYSTEM BACKUP | |
product: | |
INFORMIX SERVER / 5725A3900 / E10 - | |
Problem description: | |
Physically restoring an instance from a non-whole-system backup (no -w) will result in a system without any usable logical log files, all existing log files will have a zero unique id and no usable content, and the various dbspaces/sbspaces in the system might represent states of vastly different checkpoints, potentially many logical logs apart from each other. When now using such physically restored system for instantiating a secondary server (running 'onmode -d secondary/RSS ' on this), the list of required log files will be determined from archive checkpoint information of all the spaces in the system, with some spaces needing older logs, others only later ones. Those logs will be retrieved from primary (or log. log backup) and applied to the system, with log records from older logs pertaining to spaces only requiring later logs being skipped based on spaces' ds_logid/ds_logpos information. This skipping, though, will only work as long as the server remains online since that initial instantiation; should the secondary be restarted before reaching the highest of those ds_logid/ds_logpos LSNs, any log records older than that LSN of the space they correspond to will no longer be skipped which, as they're old and their actions long applied to the space, can lead to all kinds of log recovery failures. | |
Problem Summary: | |
**************************************************************** * USERS AFFECTED: * * Users of Informix Server prior to 14.10.xC5. * **************************************************************** * PROBLEM DESCRIPTION: * * See Error Description * **************************************************************** * RECOMMENDATION: * * Update to Informix Server 14.10.xC5. * **************************************************************** | |
Local Fix: | |
Avoid the problem by not shutting down a secondary before it reached the latest archive checkpoint of all spaces in the system. | |
Solution | |
Workaround | |
**************************************************************** * USERS AFFECTED: * * Users of Informix Server prior to 14.10.xC5. * **************************************************************** * PROBLEM DESCRIPTION: * * See Error Description * **************************************************************** * RECOMMENDATION: * * Update to Informix Server 14.10.xC5. * **************************************************************** | |
Comment | |
Fixed in Informix Server 14.10.xC5. | |
Timestamps | |
Date - problem reported : Date - problem closed : Date - last modified : | 17.09.2020 09.12.2020 09.12.2020 |
Problem solved at the following versions (IBM BugInfos) | |
Problem solved according to the fixlist(s) of the following version(s) |