Informix Error -147
-147 ISAM error: archive in progress.
The database server administrator sees this error. The action that you have requested cannot be carried out while an archive is being made. For example, you cannot add a log or a mirror during archiving. Cancel the archive or wait until it is complete, and then reissue the command.
Oninit® Troubleshooting Guidance
Reasons / Common Causes
-147 is a mutual-exclusivity rule: certain administrative operations — the official text names adding a logical log or adding a mirror specifically — can't run while an archive (backup) is actively in progress. The two touch overlapping internal structures and consistency guarantees, so the engine simply refuses to let them overlap.
- An administrative command (adding a log, adding a mirror) issued while a scheduled backup is actively running. The most straightforward case — two legitimate activities, scheduled without awareness of each other.
- Automation or scripts that don't check archive status before issuing administrative commands — a maintenance script that assumes the server is always available for its commands, run at a moment that happens to coincide with a backup window.
- Multiple administrators or teams unaware of each other's scheduled activities — one team runs a backup while another, unaware, attempts unrelated maintenance at the same time.
- A backup running far longer than expected (slow backup media, network issues, an unusually large database) extending the archive-in-progress window well past when it was planned, causing a maintenance task scheduled for "after the backup" to collide with a backup that hasn't actually finished yet.
Solutions / Resolution
- Cancel the archive if it isn't critical to complete immediately, then reissue the administrative command — a reasonable option when there's a specific, pressing need for the administrative change.
- Wait until the archive completes, then reissue the command, if there's no need to interrupt the backup.
- Coordinate scheduling between backup jobs and administrative maintenance tasks so they don't overlap — communicate maintenance windows across teams and automation explicitly.
- For automation, check archive status before issuing administrative commands, and either wait, retry with backoff, or alert with context rather than failing immediately without explanation.
- If archives are running unexpectedly long, investigate the underlying cause separately — backup media or network issues both delay legitimate backup completion and block administrative tasks for as long as the extended window lasts.
Examples
The straightforward scheduling collision
-- A nightly backup is running
onspaces -a dbspace1 -p /dev/newmirror -o 0
-- -147: cannot add a mirror while the archive is in progress
Waiting for the backup to finish, or canceling it if the mirror addition is more urgent, resolves this directly — there's no configuration change that allows the two to run concurrently.
Automation without an archive-status check
A maintenance script scheduled to add a logical log every time free log space drops below a threshold runs at 2 AM, unaware that a separate backup job is also scheduled around that time — the two occasionally collide, and the maintenance script fails with -147 whenever they do. Adding an archive-status check (and a wait-or-retry loop) to the script resolves this without needing to change either schedule.
A backup running longer than its window
A backup expected to take one hour runs for four due to slow backup media, extending well past the maintenance window planned for immediately afterward — the maintenance task fails with -147 because the archive genuinely hasn't finished, not because of a scheduling mistake. Investigating the backup media issue separately addresses the root cause, not just this one collision.
Diagnostic Checks
- Check whether an archive is currently in progress:
or the archive-specific status check for the version in use.onstat -u - Review backup job scheduling and history to confirm overlap with the attempted administrative action.
- If the archive is running unusually long, investigate backup media, network, or performance issues as a separate problem from the immediate -147.
Related Errors / Related Topics
- -100 — "ISAM error: duplicate value for a record with unique key." The other foundational ISAM-level error in this family.
- -146 — "ISAM error: the other copy of this disk is currently disabled or non-existent." A different mirroring-related administrative condition — worth knowing about together since adding a mirror is one of the two operations the official text names as blocked during an active archive.
Check archive status before assuming an administrative command should have worked — this is almost always a scheduling collision, not a configuration problem.