Adding anon v2 support to the dockerfile (#11313)

## Problem

Removed `anon` v1 support as described here:
https://github.com/neondatabase/cloud/issues/22663

Adding `anon` v2 support to re-introduce the `pg_anon` extension. 
Related Issues: https://github.com/neondatabase/cloud/issues/20456


## Summary of changes

Adding `anon` v2 support by building it in the dockerfile
This commit is contained in:
Suhas Thalanki
2025-05-01 11:22:01 -04:00
committed by GitHub
parent 5bd850d15a
commit f999632327
2 changed files with 180 additions and 0 deletions

View File

@@ -1096,6 +1096,23 @@ RUN cargo install --locked --version 0.14.1 cargo-pgrx && \
/bin/bash -c 'cargo pgrx init --pg${PG_VERSION:1}=/usr/local/pgsql/bin/pg_config'
USER root
#########################################################################################
#
# Layer "rust extensions pgrx14"
#
# Version 14 is now required by a few
# This layer should be used as a base for new pgrx extensions,
# and eventually get merged with `rust-extensions-build`
#
#########################################################################################
FROM pg-build-nonroot-with-cargo AS rust-extensions-build-pgrx14
ARG PG_VERSION
RUN cargo install --locked --version 0.14.1 cargo-pgrx && \
/bin/bash -c 'cargo pgrx init --pg${PG_VERSION:1}=/usr/local/pgsql/bin/pg_config'
USER root
#########################################################################################
#
# Layers "pg-onnx-build" and "pgrag-build"
@@ -1330,6 +1347,39 @@ COPY --from=pg_session_jwt-src /ext-src/ /ext-src/
WORKDIR /ext-src/pg_session_jwt-src
RUN cargo pgrx install --release
#########################################################################################
#
# Layer "pg-anon-pg-build"
# compile anon extension
#
#########################################################################################
FROM pg-build AS pg_anon-src
ARG PG_VERSION
COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
WORKDIR /ext-src
COPY compute/patches/anon_v2.patch .
# This is an experimental extension, never got to real production.
# !Do not remove! It can be present in shared_preload_libraries and compute will fail to start if library is not found.
ENV PATH="/usr/local/pgsql/bin/:$PATH"
RUN wget https://gitlab.com/dalibo/postgresql_anonymizer/-/archive/latest/postgresql_anonymizer-latest.tar.gz -O pg_anon.tar.gz && \
mkdir pg_anon-src && cd pg_anon-src && tar xzf ../pg_anon.tar.gz --strip-components=1 -C . && \
find /usr/local/pgsql -type f | sed 's|^/usr/local/pgsql/||' > /before.txt && \
sed -i 's/pgrx = "0.14.1"/pgrx = { version = "=0.14.1", features = [ "unsafe-postgres" ] }/g' Cargo.toml && \
patch -p1 < /ext-src/anon_v2.patch
FROM rust-extensions-build-pgrx14 AS pg-anon-pg-build
ARG PG_VERSION
COPY --from=pg_anon-src /ext-src/ /ext-src/
WORKDIR /ext-src
RUN cd pg_anon-src && \
make -j $(getconf _NPROCESSORS_ONLN) extension PG_CONFIG=/usr/local/pgsql/bin/pg_config PGVER=pg$(echo "$PG_VERSION" | sed 's/^v//') && \
make -j $(getconf _NPROCESSORS_ONLN) install PG_CONFIG=/usr/local/pgsql/bin/pg_config PGVER=pg$(echo "$PG_VERSION" | sed 's/^v//') && \
chmod -R a+r ../pg_anon-src && \
echo 'trusted = true' >> /usr/local/pgsql/share/extension/anon.control;
########################################################################################
#########################################################################################
#
# Layer "wal2json-build"
@@ -1626,6 +1676,7 @@ COPY --from=pg_uuidv7-build /usr/local/pgsql/ /usr/local/pgsql/
COPY --from=pg_roaringbitmap-build /usr/local/pgsql/ /usr/local/pgsql/
COPY --from=pg_semver-build /usr/local/pgsql/ /usr/local/pgsql/
COPY --from=wal2json-build /usr/local/pgsql /usr/local/pgsql
COPY --from=pg-anon-pg-build /usr/local/pgsql/ /usr/local/pgsql/
COPY --from=pg_ivm-build /usr/local/pgsql/ /usr/local/pgsql/
COPY --from=pg_partman-build /usr/local/pgsql/ /usr/local/pgsql/
COPY --from=pg_mooncake-build /usr/local/pgsql/ /usr/local/pgsql/

View File

@@ -0,0 +1,129 @@
diff --git a/sql/anon.sql b/sql/anon.sql
index 0cdc769..f6cc950 100644
--- a/sql/anon.sql
+++ b/sql/anon.sql
@@ -1141,3 +1141,8 @@ $$
-- TODO : https://en.wikipedia.org/wiki/L-diversity
-- TODO : https://en.wikipedia.org/wiki/T-closeness
+
+-- NEON Patches
+
+GRANT ALL ON SCHEMA anon to neon_superuser;
+GRANT ALL ON ALL TABLES IN SCHEMA anon TO neon_superuser;
diff --git a/sql/init.sql b/sql/init.sql
index 7da6553..9b6164b 100644
--- a/sql/init.sql
+++ b/sql/init.sql
@@ -74,50 +74,49 @@ $$
SECURITY LABEL FOR anon ON FUNCTION anon.load_csv IS 'UNTRUSTED';
--- load fake data from a given path
-CREATE OR REPLACE FUNCTION anon.init(
- datapath TEXT
-)
+CREATE OR REPLACE FUNCTION anon.load_fake_data()
RETURNS BOOLEAN
AS $$
DECLARE
- datapath_check TEXT;
success BOOLEAN;
+ sharedir TEXT;
+ datapath TEXT;
BEGIN
- IF anon.is_initialized() THEN
- RAISE NOTICE 'The anon extension is already initialized.';
- RETURN TRUE;
- END IF;
+ datapath := '/extension/anon/';
+ -- find the local extension directory
+ SELECT setting INTO sharedir
+ FROM pg_catalog.pg_config
+ WHERE name = 'SHAREDIR';
SELECT bool_or(results) INTO success
FROM unnest(array[
- anon.load_csv('anon.identifiers_category',datapath||'/identifiers_category.csv'),
- anon.load_csv('anon.identifier',datapath ||'/identifier.csv'),
- anon.load_csv('anon.address',datapath ||'/address.csv'),
- anon.load_csv('anon.city',datapath ||'/city.csv'),
- anon.load_csv('anon.company',datapath ||'/company.csv'),
- anon.load_csv('anon.country',datapath ||'/country.csv'),
- anon.load_csv('anon.email', datapath ||'/email.csv'),
- anon.load_csv('anon.first_name',datapath ||'/first_name.csv'),
- anon.load_csv('anon.iban',datapath ||'/iban.csv'),
- anon.load_csv('anon.last_name',datapath ||'/last_name.csv'),
- anon.load_csv('anon.postcode',datapath ||'/postcode.csv'),
- anon.load_csv('anon.siret',datapath ||'/siret.csv'),
- anon.load_csv('anon.lorem_ipsum',datapath ||'/lorem_ipsum.csv')
+ anon.load_csv('anon.identifiers_category',sharedir || datapath || '/identifiers_category.csv'),
+ anon.load_csv('anon.identifier',sharedir || datapath || '/identifier.csv'),
+ anon.load_csv('anon.address',sharedir || datapath || '/address.csv'),
+ anon.load_csv('anon.city',sharedir || datapath || '/city.csv'),
+ anon.load_csv('anon.company',sharedir || datapath || '/company.csv'),
+ anon.load_csv('anon.country',sharedir || datapath || '/country.csv'),
+ anon.load_csv('anon.email', sharedir || datapath || '/email.csv'),
+ anon.load_csv('anon.first_name',sharedir || datapath || '/first_name.csv'),
+ anon.load_csv('anon.iban',sharedir || datapath || '/iban.csv'),
+ anon.load_csv('anon.last_name',sharedir || datapath || '/last_name.csv'),
+ anon.load_csv('anon.postcode',sharedir || datapath || '/postcode.csv'),
+ anon.load_csv('anon.siret',sharedir || datapath || '/siret.csv'),
+ anon.load_csv('anon.lorem_ipsum',sharedir || datapath || '/lorem_ipsum.csv')
]) results;
RETURN success;
-
END;
$$
- LANGUAGE PLPGSQL
+ LANGUAGE plpgsql
VOLATILE
RETURNS NULL ON NULL INPUT
- PARALLEL UNSAFE -- because load_csv is unsafe
- SECURITY INVOKER
+ PARALLEL UNSAFE -- because of the EXCEPTION
+ SECURITY DEFINER
SET search_path=''
;
-SECURITY LABEL FOR anon ON FUNCTION anon.init(TEXT) IS 'UNTRUSTED';
+
+SECURITY LABEL FOR anon ON FUNCTION anon.load_fake_data IS 'UNTRUSTED';
-- People tend to forget the anon.init() step
-- This is a friendly notice for them
@@ -144,7 +143,7 @@ SECURITY LABEL FOR anon ON FUNCTION anon.notice_if_not_init IS 'UNTRUSTED';
CREATE OR REPLACE FUNCTION anon.load(TEXT)
RETURNS BOOLEAN AS
$$
- SELECT anon.init($1);
+ SELECT anon.init();
$$
LANGUAGE SQL
VOLATILE
@@ -159,16 +158,16 @@ SECURITY LABEL FOR anon ON FUNCTION anon.load(TEXT) IS 'UNTRUSTED';
CREATE OR REPLACE FUNCTION anon.init()
RETURNS BOOLEAN
AS $$
- WITH conf AS (
- -- find the local extension directory
- SELECT setting AS sharedir
- FROM pg_catalog.pg_config
- WHERE name = 'SHAREDIR'
- )
- SELECT anon.init(conf.sharedir || '/extension/anon/')
- FROM conf;
+BEGIN
+ IF anon.is_initialized() THEN
+ RAISE NOTICE 'The anon extension is already initialized.';
+ RETURN TRUE;
+ END IF;
+
+ RETURN anon.load_fake_data();
+END;
$$
- LANGUAGE SQL
+ LANGUAGE plpgsql
VOLATILE
PARALLEL UNSAFE -- because init is unsafe
SECURITY INVOKER