DBSync

Important

An average pool operator may not require cardano-db-sync at all. Please verify if it is required for your use as mentioned here.

Ensure the Pre-Requisites are in place before you proceed.
The Cardano DB Sync relies on an existing PostgreSQL server. To keep the focus on building dbsync tool, and not how to setup postgres itself, you can refer to Sample Local PostgreSQL Server Deployment instructions for setting up a Postgres instance. Specifically, we expect the PGPASSFILE environment variable is set as per the instructions in the sample guide, for db-sync to be able to connect.
One of the biggest obstacles for user experience when running dbsync is ensuring you satisfy EACH of the points mentioned in System Requirements here. Also, note that we do not advise running dbsync on mainnet if your RAM is below 48GB.

Build Instructions⚓︎

Clone the repository⚓︎

Execute the below to clone the cardano-db-sync repository to $HOME/git folder on your system:

cd ~/git
git clone https://github.com/intersectmbo/cardano-db-sync
cd cardano-db-sync

Build Cardano DB Sync⚓︎

You can use the instructions below to build the latest release of cardano-db-sync.

git fetch --tags --all
git pull
# Include the cardano-crypto-praos and libsodium components for db-sync
# On CentOS 7 (GCC 4.8.5) we should also do
# echo -e "package cryptonite\n  flags: -use_target_attributes" >> cabal.project.local
# Replace tag against checkout if you do not want to build the latest released version
git checkout $(curl -sLf https://api.github.com/repos/intersectmbo/cardano-db-sync/releases/latest | jq -r .tag_name)
# Use `-l` argument if you'd like to use system libsodium instead of IOG fork of libsodium while compiling
$CNODE_HOME/scripts/cabal-build-all.sh

The above would copy the cardano-db-sync binary into ~/.local/bin folder.

Prepare DB for sync⚓︎

Now that binaries are available, let's create our database (when going through breaking changes, you may need to use --recreatedb instead of --createdb used for the first time. Again, we expect that PGPASSFILE environment variable is already set (refer to the top of this guide for sample instructions):

cd ~/git/cardano-db-sync
# scripts/postgresql-setup.sh --dropdb #if exists already, will fail if it doesnt - thats OK
scripts/postgresql-setup.sh --createdb
# Password:
# Password:
# All good!

Verify you can see "All good!" as above!

Create Symlink to schema folder⚓︎

DBSync instance requires the schema files from the git repository to be present and available to the dbsync instance. You can either clone the ~/git/cardano-db-sync/schema folder OR create a symlink to the folder and make it available to the startup command we will be using. We will use the latter in sample below:

ln -s ~/git/cardano-db-sync/schema $CNODE_HOME/guild-db/schema

Restore using Snapshot⚓︎

If you're running a mainnet/preview/preprod instance of dbsync, you might want to consider use of dbsync snapshots as documented here. The snapshot files as of recent epoch are available via links in release notes.

At high-level, this would involve steps as below (read and update paths as per your environment):

# Replace the actual link below with the latest one from release notes
wget https://update-cardano-mainnet.iohk.io/cardano-db-sync/13/db-sync-snapshot-schema-13-block-7622755-x86_64.tgz
rm -rf ${CNODE_HOME}/guild-db/ledger-state ; mkdir -p ${CNODE_HOME}/guild-db/ledger-state
cd -; cd ~/git/cardano-db-sync
scripts/postgresql-setup.sh --restore-snapshot /tmp/dbsyncsnap.tgz ${CNODE_HOME}/guild-db/ledger-state
# The restore may take a while, please be patient and do not interrupt the restore process. Once restore is successful, you may delete the downloaded snapshot as below:
#   rm -f /tmp/dbsyncsnap.tgz

Test running dbsync manually at terminal⚓︎

In order to verify that you can run dbsync, before making a start - you'd want to ensure that you can run it interactively once. To do so, try the commands below:

cd $CNODE_HOME/scripts
export PGPASSFILE=$CNODE_HOME/priv/.pgpass
./dbsync.sh

You can monitor logs if needed via parallel session using tail -10f $CNODE_HOME/logs/dbsync.json. If there are no error, you would want to press Ctrl-C to stop the dbsync.sh execution and deploy it as a systemd service. To do so, use the commands below (the creation of file is done using sudo permissions, but you can always deploy it manually):

cd $CNODE_HOME/scripts
./dbsync.sh -d
# Deploying cnode-dbsync.service as systemd service..
# cnode-dbsync.service deployed successfully!!

Now to start dbsync instance, you can run sudo systemctl start cnode-dbsync

Note

Note that dbsync while syncs, it might defer creation of indexes/constraints to speed up initial catch up. Once relatively closer to tip, this will initiate creation of indexes - which can take a while in background. Thus, you might notice the query timings right after reaching to tip might not be as good.

Update DBSync⚓︎

Updating dbsync can have different tasks depending on the versions involved. We attempt to briefly explain the tasks involved:

Shutdown dbsync (eg: sudo systemctl stop cnode-dbsync)
Update binaries (either download pre-compiled binaries via guild-deploy.sh or using build instructions above)
Go to your git folder, pull and checkout to latest version as in example below (if you were to switch to 13.1.1.3):
```
cd ~/git/cardano-db-sync
git pull
git checkout 13.1.1.3
```
If going through major version update (eg: 13.x.x.x to 14.x.x.x), you might need to rebuild and resync db from scratch, you may still follow the section to restore using snapshot to save some time (as long as you use a compatible snapshot).
If the underlying cardano-node version has changed (specifically if it's ledger-state schema is different), you'd also need to clear the ledger-state directory (eg: rm -rf $CNODE_HOME/guild-db/ledger-state)
Test that dbsync.sh starts up fine manually as described above. If it does, stop it and go ahead with startup of systemd service (i.e. sudo systemctl start cnode-dbsync)

Validation⚓︎

To validate, connect to your postgres instance and execute commands as per below:

export PGPASSFILE=$CNODE_HOME/priv/.pgpass
psql cexplorer

You should be at the psql prompt, you can check the tables and verify they're populated:

\dt
select * from meta;

A sample output of the above two commands may look like below (the number of tables and names may vary between versions):

cexplorer=# \dt
List of relations
 Schema |           Name            | Type  | Owner
--------+---------------------------+-------+-------
 public | ada_pots                  | table | centos
 public | admin_user                | table | centos
 public | block                     | table | centos
 public | delegation                | table | centos
 public | delisted_pool             | table | centos
 public | epoch                     | table | centos
 public | epoch_param               | table | centos
 public | epoch_stake               | table | centos
 public | ma_tx_mint                | table | centos
 public | ma_tx_out                 | table | centos
 public | meta                      | table | centos
 public | orphaned_reward           | table | centos
 public | param_proposal            | table | centos
 public | pool_hash                 | table | centos
 public | pool_meta_data            | table | centos
 public | pool_metadata             | table | centos
 public | pool_metadata_fetch_error | table | centos
 public | pool_metadata_ref         | table | centos
 public | pool_owner                | table | centos
 public | pool_relay                | table | centos
 public | pool_retire               | table | centos
 public | pool_update               | table | centos
 public | pot_transfer              | table | centos
 public | reserve                   | table | centos
 public | reserved_ticker           | table | centos
 public | reward                    | table | centos
 public | schema_version            | table | centos
 public | slot_leader               | table | centos
 public | stake_address             | table | centos
 public | stake_deregistration      | table | centos
 public | stake_registration        | table | centos
 public | treasury                  | table | centos
 public | tx                        | table | centos
 public | tx_in                     | table | centos
 public | tx_metadata               | table | centos
 public | tx_out                    | table | centos
 public | withdrawal                | table | centos
(37 rows)



select * from meta;
 id |     start_time      | network_name
----+---------------------+--------------
  1 | 2017-09-23 21:44:51 | mainnet
(1 row)