Well, the time is nigh. Our SQLite database has hit 1.3GB and creates a
900MB dump file, which is big enough to cause our bacula Xen virtual
machine to run out of space when running a catalog backup.
We have a Postgres (7.4) server with plenty of space, so I thought I'd
try importing the catalog into that.. things didn't go so smoothly, so I
ended up knocking up a script to do the import.
I haven't actually finished it yet, and I know there are some
optimisations which could be made, but I hope this is useful to someone:
What could be better:
* Automatic reorganisation of CREATE statements so that they execute in
the correct order
* Commandline parameters
* Handling of arbitrary conversions (Postgres -> MySQL, MySQL ->
Postgres, SQLite -> MySQL, MySQL -> SQLite, Postgres -> SQLite)
Note that it would be a good idea to turn off the fsync stuff on
Postgres while you do the import - I haven't benchmarked, but it should
make things go a good bit quicker.
I haven't actually tested this yet, so it's probably broken and riddled
with bugs. Still, enjoy :)
--
Russell Howe
[EMAIL PROTECTED]
#!/bin/bash
# Import an SQLite dump of a Bacula catalog into Postgres
# Designed for v1.63.3 (as found on Debian sarge)
#
# v0.5
#
# Copyright (c) 2006 Russell Howe <[EMAIL PROTECTED]>
# Permission is hereby granted, free of charge, to any person obtaining a
# copy of this software and associated documentation files (the "Software"),
# to deal in the Software without restriction, including without limitation
# the rights to use, copy, modify, merge, publish, distribute, sublicense,
# and/or sell copies of the Software, and to permit persons to whom the
# Software is furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice shall be included in
# all copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
# OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
# FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
# DEALINGS IN THE SOFTWARE.
FILE=bacula.sql
# Tables, in order of size
TABLES=(File Filename Path Job Media Pool CDImages Counters Version Client
FileSet JobMedia NextId UnsavedFiles BaseFiles)
# Tables, in insert order
TABLESINSERT=(Pool CDImages Client Counters FileSet Filename Job Media JobMedia
NextId Path File UnsavedFiles Version BaseFiles)
DBNAME=bacula
LOGFILE="/var/tmp/sqlite2pgsql.$$.log"
importdata() {
if [ "x" == "x$1" ]; then
echo "Error: importdata() called without an argument. Aborting."
exit 1
fi
SQLFILE="$1"
if [ ! -r "$SQLFILE" ]; then
echo "Error: Cannot read from $SQLFILE. Aborting."
exit 1
fi
echo -n "Loading $SQLFILE into database $DBNAME..."
psql -d "$DBNAME" -f "$SQLFILE" || (
echo "Failed to load $SQLFILE into database $DBNAME. psql
exited with return code $?. Aborting."
exit 1
)
}
# Go through each of the table names, splitting the INSERT statements off
# into seperate files
for table in [EMAIL PROTECTED]; do
SRC="$FILE.other"
if [ ! -f "$FILE.other" ]; then
SRC="$FILE"
fi
PATTERN="^INSERT INTO $table "
if [ ! -f "$FILE.data.$table" ]; then
echo -n "Separating $table table from database dump..."
echo "BEGIN TRANSACTION" > "$FILE.data.$table.tmp"
grep "$PATTERN" "$SRC" >> "$FILE.data.$table.tmp"
echo "COMMIT" >> "$FILE.data.$table.tmp"
mv "$FILE.data.$table.tmp" "$FILE.data.$table"
echo "done. ($FILE.data.$table)"
echo -n "Stripping matched lines from the source file to speed
up the next round..."
grep -v "$PATTERN" "$SRC" > "$FILE.other.tmp"
mv "$FILE.other.tmp" "$FILE.other"
echo "done."
else
echo "$FILE.data.$table already exists. Assuming this table has
already been split"
echo "off from the main dump. Not regenerating."
fi
done
echo "Seperating DDL statements from INSERT statements"
grep -v "^INSERT" "$FILE.other" > "$FILE.ddl"
echo "DDL statements are now in $FILE.ddl"
grep "^INSERT" "$FILE.other" > "$FILE.data.other"
echo "Any remaining INSERT statements are now in $FILE.data.other"
echo "Fixing up datatypes used in the DDL..."
sed -e 's/TINYINT/SMALLINT/g' \
-e 's/DATETIME/TIMESTAMP/g' \
-e 's/INTEGER UNSIGNED/INTEGER/g' \
-e 's/BIGINT UNSIGNED/BIGINT/g' \
-e 's/INTEGER AUTOINCREMENT/SERIAL/g' \
-e s/\ DEFAULT\ \"\"/\ DEFAULT\ \'\'/g \
-e s#\ TIMESTAMP\ DEFAULT\ 0#\ TIMESTAMP\ DEFAULT\ \'1/1/1970\'#g
"$FILE.ddl" > "$FILE.ddl.postgres"
echo "Fixing Pool table..."
sed -e 's/,0,0);$/,NULL,NULL);/' "$FILE.data.Pool" > "$FILE.data.Pool.fixed"
echo "OK, we should be ready to import data into PostgreSQL now. DDL first..."
echo "This will probably fail the first time. You will have to edit $FILE.other"
echo "and rearrange the CREATE TABLE statements so that the tables are created"
echo "in the correct order."
echo "After editing $FILE.other, simply rerun this script and it will carry on"
echo "where it left off."
importdata "$FILE.ddl.postgres"
for table in [EMAIL PROTECTED] other; do
IMPORTFILE="$FILE.data.$table"
if [ -f "$FILE.data.$table.fixed" ]; then
IMPORTFILE="$FILE.data.$table.fixed"
fi
importdata "$IMPORTFILE" 2>&1 |tee -a "$LOGFILE"
done
echo "All done! Check $LOGFILE for errors."
