Tech Preview Feature: Added option to skip regeneration of offline_db.
By default, after creating a checkpoint from the offline_db, the
daily_checkpoint.sh script deletes the db.* files in the offline_db,
and then recreates them from from the checkpoint. This has several
benefits, including verifying the usability of the checkpoint and
replacing potentially bloated databases with more compact, freshly
rebuilt from a checkpoint databases that can be taken advantage of
when refresh_P4ROOT_from_offline_db.sh is used.
However, in some circumstances it may be better to forego those
benefits in favor of reducing the duration of the daily_checkpoint.sh
script, and maintaining the usability of db.* files in the offline_db
for a fast recovery. This is especially the case where checkpoint
operations take several hours.
This initial implementation is as a Tech Preview feature, sans
documentation. A better follow-on implementation may be tied to
SDP-530 and SDP-568 (rewriting command line usage and logging).
This feature can be enabled with a command like the following, with
this example being for SDP instance 1:
echo 'export SDP_DAILY_OFFLINE_DB_REGEN=0' >> /p4/common/site/config/p4_1.vars.local