dump_instance.sh and load_instance.sh scripts

Issue #65 - Prepare regression test covering database backup/restore

Issue #4 use database backup to restore HAF

The scripts save or load the haf database along with the hived state.

OPTIONS:
  --backup-dir=DIR                 Allows to specify where the dumps are located, required
  --hived-executable-path=PATH     Allows to specify where the main executable is located, required
  --hived-data-dir=DIR             Allows to specify where node data is located, required
  --haf-db-name=NAME               Allows to specify a name of database to store HAF data, required
  --haf-db-admin=NAME              Allows to specify a name of database admin role having permission to create the database
  --haf-db-host=VALUE              Allows to specify a PostgreSQL host location
  --haf-db-port=NUMBER             Allows to specify a PostgreSQL operating port
  --haf-db-jobs=NUMBER             Allows to specify how many processes are run during dump/restore
  --override-existing-backup-dir   Erases backup directory and its contents
  --help                           Display this help screen and exit

A note using Docker: hivemind SELECT * on hivemind_app view issues an error "could not resize shared memory segment" . Fixed by adding --shm-size=2gb to the docker run command.

tests/integration/system/haf/test_pg_dump.py

@@ -11,7 +11,7 @@ from local_tools import make_fork, wait_for_irreversible_progress, run_networks,
 
 
 # MTTK TODO once the errrors disappear we could use single pg_restore command
-ERRRORS_IN_CREATE_POLICY = True
+ERRORS_IN_CREATE_POLICY = True
 
 
 
@@ -85,7 +85,7 @@ def wipe_db(db_name):
 
 
 def pg_restore(target_db_name):
-    if ERRRORS_IN_CREATE_POLICY:
+    if ERRORS_IN_CREATE_POLICY:
         # restore pre-data
         shell(f"pg_restore --section=pre-data  -Fc -d {target_db_name}   adump.Fcsql")