Warn about unusual paths in timescaledb-ha image (#858)

Co-authored-by: Vladimir Bobrikov <[email protected]>
Co-authored-by: Lev Gorodetskiy <[email protected]>

Author: 3 people

CHANGELOG.md

@@ -12,6 +12,8 @@ The format is based on [Keep a Changelog], and this project adheres to [Semantic
 
 ### Fixed
 
+
+- cli: Added warning for timescaledb-ha users
 - demos: Fixed decimal overflow in `demo_uniswap` project.
 - evm.node: Fixed incorrect log request parameters.
 - evm.subsquid.events: Fixed issue with determining the last level when syncing with node.

docs/6.deployment/1.database.md

@@ -24,3 +24,30 @@ Immune tables support for SQLite is experimental and requires `advanced.unsafe_s
 Latest PostgreSQL and TimescaleDB versions are recommended due to significant performance improvements in recent releases (see [Feature matrix](https://www.postgresql.org/about/featurematrix/) page).
 
 Usually it's okay to use different database engines for development and production, but be careful with SQL scripts and column types that can behave differently.
+
+## TimescaleDB
+
+TimescaleDB is a PostgreSQL extension that provides time-series data management capabilities. You can use it with DipDup as a drop-in replacement for PostgreSQL.
+
+You can choose `timescale/timescaledb` or `timescale/timescaledb-ha` Docker images when initializing DipDup project.
+
+::banner{type="warning"}
+Be careful! Unlike other PostgreSQL images `timescale/timescaledb-ha` uses `/home/postgres/pgdata/data` as a persistent volume.
+::
+
+Use SQL scripts in `on_reindex` directory to prepare the database. First, create a hypertable replacing primary key with composite one:
+
+```sql [sql/on_reindex/00_prepare_db.sql]
+{{ #include ../src/demo_uniswap/sql/on_reindex/00_prepare_db.sql }}
+```
+
+Now you can create a continuous aggregate on top of the hypertable:
+
+```sql [sql/on_reindex/01_create_mv_token_price.sql]
+{{ #include ../src/demo_uniswap/sql/on_reindex/01_create_mv_token_price.sql }}
+```
+
+For more information visit the official TimescaleDB documentation:
+
+- [Hypertables](https://docs.timescale.com/use-timescale/latest/hypertables/)
+- [Continuous aggregates](https://docs.timescale.com/use-timescale/continuous-aggregates/)

src/dipdup/project.py

@@ -214,8 +214,9 @@ def answers_from_terminal() -> Answers:
         ),
         default=0,
     )
-    if 'timescale-ha' in answers['postgres_image']:
+    if 'timescaledb-ha' in answers['postgres_image']:
         answers['postgres_data_path'] = '/home/postgres/pgdata/data'
+        echo('`timescaledb-ha` Docker image uses `/home/postgres/pgdata/data` as a data path; generated files were updated accordingly.', fg='yellow')
 
     big_yellow_echo('Miscellaneous tunables; leave default values if unsure')