Generate separate snapshots for logical databases (#13792)

* Generate separate snapshots for sqlite, postgres and common
* Cleanup postgres dbs in the TRAP
* Say which logical DB we're applying updates to
* Run background updates on the state DB
* Add new option for accepting a SCHEMA_NUMBER
This commit is contained in:
David Robertson 2022-09-20 14:14:12 +01:00 committed by GitHub
parent 42d261c32f
commit fff9b955fa
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 140 additions and 46 deletions

1
changelog.d/13792.misc Normal file
View File

@ -0,0 +1 @@
Update the script which makes full schema dumps.

View File

@ -2,23 +2,16 @@
# #
# This script generates SQL files for creating a brand new Synapse DB with the latest # This script generates SQL files for creating a brand new Synapse DB with the latest
# schema, on both SQLite3 and Postgres. # schema, on both SQLite3 and Postgres.
#
# It does so by having Synapse generate an up-to-date SQLite DB, then running
# synapse_port_db to convert it to Postgres. It then dumps the contents of both.
export PGHOST="localhost" export PGHOST="localhost"
POSTGRES_DB_NAME="synapse_full_schema.$$" POSTGRES_MAIN_DB_NAME="synapse_full_schema_main.$$"
POSTGRES_COMMON_DB_NAME="synapse_full_schema_common.$$"
SQLITE_SCHEMA_FILE="schema.sql.sqlite" POSTGRES_STATE_DB_NAME="synapse_full_schema_state.$$"
SQLITE_ROWS_FILE="rows.sql.sqlite"
POSTGRES_SCHEMA_FILE="full.sql.postgres"
POSTGRES_ROWS_FILE="rows.sql.postgres"
REQUIRED_DEPS=("matrix-synapse" "psycopg2") REQUIRED_DEPS=("matrix-synapse" "psycopg2")
usage() { usage() {
echo echo
echo "Usage: $0 -p <postgres_username> -o <path> [-c] [-n] [-h]" echo "Usage: $0 -p <postgres_username> -o <path> [-c] [-n <schema number>] [-h]"
echo echo
echo "-p <postgres_username>" echo "-p <postgres_username>"
echo " Username to connect to local postgres instance. The password will be requested" echo " Username to connect to local postgres instance. The password will be requested"
@ -27,11 +20,16 @@ usage() {
echo " CI mode. Prints every command that the script runs." echo " CI mode. Prints every command that the script runs."
echo "-o <path>" echo "-o <path>"
echo " Directory to output full schema files to." echo " Directory to output full schema files to."
echo "-n <schema number>"
echo " Schema number for the new snapshot. Used to set the location of files within "
echo " the output directory, mimicking that of synapse/storage/schemas."
echo " Defaults to 9999."
echo "-h" echo "-h"
echo " Display this help text." echo " Display this help text."
} }
while getopts "p:co:h" opt; do SCHEMA_NUMBER="9999"
while getopts "p:co:hn:" opt; do
case $opt in case $opt in
p) p)
export PGUSER=$OPTARG export PGUSER=$OPTARG
@ -48,6 +46,9 @@ while getopts "p:co:h" opt; do
usage usage
exit exit
;; ;;
n)
SCHEMA_NUMBER="$OPTARG"
;;
\?) \?)
echo "ERROR: Invalid option: -$OPTARG" >&2 echo "ERROR: Invalid option: -$OPTARG" >&2
usage usage
@ -95,12 +96,21 @@ cd "$(dirname "$0")/.."
TMPDIR=$(mktemp -d) TMPDIR=$(mktemp -d)
KEY_FILE=$TMPDIR/test.signing.key # default Synapse signing key path KEY_FILE=$TMPDIR/test.signing.key # default Synapse signing key path
SQLITE_CONFIG=$TMPDIR/sqlite.conf SQLITE_CONFIG=$TMPDIR/sqlite.conf
SQLITE_DB=$TMPDIR/homeserver.db SQLITE_MAIN_DB=$TMPDIR/main.db
SQLITE_STATE_DB=$TMPDIR/state.db
SQLITE_COMMON_DB=$TMPDIR/common.db
POSTGRES_CONFIG=$TMPDIR/postgres.conf POSTGRES_CONFIG=$TMPDIR/postgres.conf
# Ensure these files are delete on script exit # Ensure these files are delete on script exit
# TODO: the trap should also drop the temp postgres DB cleanup() {
trap 'rm -rf $TMPDIR' EXIT echo "Cleaning up temporary sqlite database and config files..."
rm -r "$TMPDIR"
echo "Cleaning up temporary Postgres database..."
dropdb --if-exists "$POSTGRES_COMMON_DB_NAME"
dropdb --if-exists "$POSTGRES_MAIN_DB_NAME"
dropdb --if-exists "$POSTGRES_STATE_DB_NAME"
}
trap 'cleanup' EXIT
cat > "$SQLITE_CONFIG" <<EOF cat > "$SQLITE_CONFIG" <<EOF
server_name: "test" server_name: "test"
@ -110,10 +120,22 @@ macaroon_secret_key: "abcde"
report_stats: false report_stats: false
database: databases:
common:
name: "sqlite3" name: "sqlite3"
data_stores: []
args: args:
database: "$SQLITE_DB" database: "$SQLITE_COMMON_DB"
main:
name: "sqlite3"
data_stores: ["main"]
args:
database: "$SQLITE_MAIN_DB"
state:
name: "sqlite3"
data_stores: ["state"]
args:
database: "$SQLITE_STATE_DB"
# Suppress the key server warning. # Suppress the key server warning.
trusted_key_servers: [] trusted_key_servers: []
@ -127,13 +149,32 @@ macaroon_secret_key: "abcde"
report_stats: false report_stats: false
database: databases:
common:
name: "psycopg2" name: "psycopg2"
data_stores: []
args: args:
user: "$PGUSER" user: "$PGUSER"
host: "$PGHOST" host: "$PGHOST"
password: "$PGPASSWORD" password: "$PGPASSWORD"
database: "$POSTGRES_DB_NAME" database: "$POSTGRES_COMMON_DB_NAME"
main:
name: "psycopg2"
data_stores: ["main"]
args:
user: "$PGUSER"
host: "$PGHOST"
password: "$PGPASSWORD"
database: "$POSTGRES_MAIN_DB_NAME"
state:
name: "psycopg2"
data_stores: ["state"]
args:
user: "$PGUSER"
host: "$PGHOST"
password: "$PGPASSWORD"
database: "$POSTGRES_STATE_DB_NAME"
# Suppress the key server warning. # Suppress the key server warning.
trusted_key_servers: [] trusted_key_servers: []
@ -148,33 +189,76 @@ echo "Running db background jobs..."
synapse/_scripts/update_synapse_database.py --database-config "$SQLITE_CONFIG" --run-background-updates synapse/_scripts/update_synapse_database.py --database-config "$SQLITE_CONFIG" --run-background-updates
# Create the PostgreSQL database. # Create the PostgreSQL database.
echo "Creating postgres database..." echo "Creating postgres databases..."
createdb --lc-collate=C --lc-ctype=C --template=template0 "$POSTGRES_DB_NAME" createdb --lc-collate=C --lc-ctype=C --template=template0 "$POSTGRES_COMMON_DB_NAME"
createdb --lc-collate=C --lc-ctype=C --template=template0 "$POSTGRES_MAIN_DB_NAME"
createdb --lc-collate=C --lc-ctype=C --template=template0 "$POSTGRES_STATE_DB_NAME"
echo "Running db background jobs..." echo "Running db background jobs..."
synapse/_scripts/update_synapse_database.py --database-config "$POSTGRES_CONFIG" --run-background-updates synapse/_scripts/update_synapse_database.py --database-config "$POSTGRES_CONFIG" --run-background-updates
# Delete schema_version, applied_schema_deltas and applied_module_schemas tables
# Also delete any shadow tables from fts4
echo "Dropping unwanted db tables..." echo "Dropping unwanted db tables..."
SQL="
# Some common tables are created and updated by Synapse itself and do not belong in the
# schema.
DROP_APP_MANAGED_TABLES="
DROP TABLE schema_version; DROP TABLE schema_version;
DROP TABLE schema_compat_version;
DROP TABLE applied_schema_deltas; DROP TABLE applied_schema_deltas;
DROP TABLE applied_module_schemas; DROP TABLE applied_module_schemas;
" "
sqlite3 "$SQLITE_DB" <<< "$SQL" # Other common tables are not created by Synapse and do belong in the schema.
psql "$POSTGRES_DB_NAME" -w <<< "$SQL" # TODO: we could derive DROP_COMMON_TABLES from the dump of the common-only DB. But
# since there's only one table there, I haven't bothered to do so.
DROP_COMMON_TABLES="$DROP_APP_MANAGED_TABLES
DROP TABLE background_updates;
"
echo "Dumping SQLite3 schema to '$OUTPUT_DIR/$SQLITE_SCHEMA_FILE' and '$OUTPUT_DIR/$SQLITE_ROWS_FILE'..." sqlite3 "$SQLITE_COMMON_DB" <<< "$DROP_APP_MANAGED_TABLES"
sqlite3 "$SQLITE_DB" ".schema --indent" > "$OUTPUT_DIR/$SQLITE_SCHEMA_FILE" sqlite3 "$SQLITE_MAIN_DB" <<< "$DROP_COMMON_TABLES"
sqlite3 "$SQLITE_DB" ".dump --data-only --nosys" > "$OUTPUT_DIR/$SQLITE_ROWS_FILE" sqlite3 "$SQLITE_STATE_DB" <<< "$DROP_COMMON_TABLES"
psql "$POSTGRES_COMMON_DB_NAME" -w <<< "$DROP_APP_MANAGED_TABLES"
psql "$POSTGRES_MAIN_DB_NAME" -w <<< "$DROP_COMMON_TABLES"
psql "$POSTGRES_STATE_DB_NAME" -w <<< "$DROP_COMMON_TABLES"
echo "Dumping Postgres schema to '$OUTPUT_DIR/$POSTGRES_SCHEMA_FILE' and '$OUTPUT_DIR/$POSTGRES_ROWS_FILE'..." # For Reasons(TM), SQLite's `.schema` also dumps out "shadow tables", the implementation
pg_dump --format=plain --schema-only --no-tablespaces --no-acl --no-owner "$POSTGRES_DB_NAME" | sed -e '/^$/d' -e '/^--/d' -e 's/public\.//g' -e '/^SET /d' -e '/^SELECT /d' > "$OUTPUT_DIR/$POSTGRES_SCHEMA_FILE" # details behind full text search tables. Omit these from the dumps.
pg_dump --format=plain --data-only --inserts --no-tablespaces --no-acl --no-owner "$POSTGRES_DB_NAME" | sed -e '/^$/d' -e '/^--/d' -e 's/public\.//g' -e '/^SET /d' -e '/^SELECT /d' > "$OUTPUT_DIR/$POSTGRES_ROWS_FILE"
echo "Cleaning up temporary Postgres database..." sqlite3 "$SQLITE_MAIN_DB" <<< "
dropdb $POSTGRES_DB_NAME DROP TABLE event_search_content;
DROP TABLE event_search_segments;
DROP TABLE event_search_segdir;
DROP TABLE event_search_docsize;
DROP TABLE event_search_stat;
DROP TABLE user_directory_search_content;
DROP TABLE user_directory_search_segments;
DROP TABLE user_directory_search_segdir;
DROP TABLE user_directory_search_docsize;
DROP TABLE user_directory_search_stat;
"
echo "Dumping SQLite3 schema..."
mkdir -p "$OUTPUT_DIR/"{common,main,state}"/full_schema/$SCHEMA_NUMBER"
sqlite3 "$SQLITE_COMMON_DB" ".schema --indent" > "$OUTPUT_DIR/common/full_schema/$SCHEMA_NUMBER/full.sql.sqlite"
sqlite3 "$SQLITE_COMMON_DB" ".dump --data-only --nosys" >> "$OUTPUT_DIR/common/full_schema/$SCHEMA_NUMBER/full.sql.sqlite"
sqlite3 "$SQLITE_MAIN_DB" ".schema --indent" > "$OUTPUT_DIR/main/full_schema/$SCHEMA_NUMBER/full.sql.sqlite"
sqlite3 "$SQLITE_MAIN_DB" ".dump --data-only --nosys" >> "$OUTPUT_DIR/main/full_schema/$SCHEMA_NUMBER/full.sql.sqlite"
sqlite3 "$SQLITE_STATE_DB" ".schema --indent" > "$OUTPUT_DIR/state/full_schema/$SCHEMA_NUMBER/full.sql.sqlite"
sqlite3 "$SQLITE_STATE_DB" ".dump --data-only --nosys" >> "$OUTPUT_DIR/state/full_schema/$SCHEMA_NUMBER/full.sql.sqlite"
cleanup_pg_schema() {
sed -e '/^$/d' -e '/^--/d' -e 's/public\.//g' -e '/^SET /d' -e '/^SELECT /d'
}
echo "Dumping Postgres schema..."
pg_dump --format=plain --schema-only --no-tablespaces --no-acl --no-owner "$POSTGRES_COMMON_DB_NAME" | cleanup_pg_schema > "$OUTPUT_DIR/common/full_schema/$SCHEMA_NUMBER/full.sql.postgres"
pg_dump --format=plain --data-only --inserts --no-tablespaces --no-acl --no-owner "$POSTGRES_COMMON_DB_NAME" | cleanup_pg_schema >> "$OUTPUT_DIR/common/full_schema/$SCHEMA_NUMBER/full.sql.postgres"
pg_dump --format=plain --schema-only --no-tablespaces --no-acl --no-owner "$POSTGRES_MAIN_DB_NAME" | cleanup_pg_schema > "$OUTPUT_DIR/main/full_schema/$SCHEMA_NUMBER/full.sql.postgres"
pg_dump --format=plain --data-only --inserts --no-tablespaces --no-acl --no-owner "$POSTGRES_MAIN_DB_NAME" | cleanup_pg_schema >> "$OUTPUT_DIR/main/full_schema/$SCHEMA_NUMBER/full.sql.postgres"
pg_dump --format=plain --schema-only --no-tablespaces --no-acl --no-owner "$POSTGRES_STATE_DB_NAME" | cleanup_pg_schema > "$OUTPUT_DIR/state/full_schema/$SCHEMA_NUMBER/full.sql.postgres"
pg_dump --format=plain --data-only --inserts --no-tablespaces --no-acl --no-owner "$POSTGRES_STATE_DB_NAME" | cleanup_pg_schema >> "$OUTPUT_DIR/state/full_schema/$SCHEMA_NUMBER/full.sql.postgres"
echo "Done! Files dumped to: $OUTPUT_DIR" echo "Done! Files dumped to: $OUTPUT_DIR"

View File

@ -48,10 +48,13 @@ class MockHomeserver(HomeServer):
def run_background_updates(hs: HomeServer) -> None: def run_background_updates(hs: HomeServer) -> None:
store = hs.get_datastores().main main = hs.get_datastores().main
state = hs.get_datastores().state
async def run_background_updates() -> None: async def run_background_updates() -> None:
await store.db_pool.updates.run_background_updates(sleep=False) await main.db_pool.updates.run_background_updates(sleep=False)
if state:
await state.db_pool.updates.run_background_updates(sleep=False)
# Stop the reactor to exit the script once every background update is run. # Stop the reactor to exit the script once every background update is run.
reactor.stop() reactor.stop()
@ -97,8 +100,11 @@ def main() -> None:
# Load, process and sanity-check the config. # Load, process and sanity-check the config.
hs_config = yaml.safe_load(args.database_config) hs_config = yaml.safe_load(args.database_config)
if "database" not in hs_config: if "database" not in hs_config and "databases" not in hs_config:
sys.stderr.write("The configuration file must have a 'database' section.\n") sys.stderr.write(
"The configuration file must have a 'database' or 'databases' section. "
"See https://matrix-org.github.io/synapse/latest/usage/configuration/config_documentation.html#database"
)
sys.exit(4) sys.exit(4)
config = HomeServerConfig() config = HomeServerConfig()

View File

@ -285,7 +285,10 @@ class BackgroundUpdater:
back_to_back_failures = 0 back_to_back_failures = 0
try: try:
logger.info("Starting background schema updates") logger.info(
"Starting background schema updates for database %s",
self._database_name,
)
while self.enabled: while self.enabled:
try: try:
result = await self.do_next_background_update(sleep) result = await self.do_next_background_update(sleep)