Five minutes after my data warehouse project's weekly ETL script for
its logged database completed, an assert failure occurred and the
informixserver halted at a checkpoint. When I forced it to continue,
the instance remained online but an entire dbspace with constituent
chunks was marked down. The dbspace contains data from a single table
that is fragmented across multiple dbspaces. If I execute an onbar
restore for the dbspace, will the data be reliably restored despite
the fragmentation or any referential integrity constraints?

Using IDS 10.00.FC7 on RHELAS9/2.6.9.

Yes ... you may have to play with backing up the logs(obviously using onbar) first.

Question is more "why did it af?". If due to a bad disk (i.e. not an Informix defect) then you should sort that out, otherwise why will it not happen again?

JJ