wallet: Store transactions in a separate sqlite table

achow101 commented at 10:52 pm on July 21, 2025: member

The wallet uses SQLite as a key-value store even though SQLite is a powerful relational database engine. This causes us numerous headaches due to the need to serialize multiple fields together, and since record read from the database during loading can come in any order.

Notably, for transactions, if we were able to load them in the transaction order assigned by the wallet, PRs like #27865 would be a bit simpler and easier to reason about.

This PR makes it possible for us to do that by storing transactions in a separate table. This table has a column for each field in CWalletTx, which lets us use a SQL statement like SELECT * FROM transactions ORDER BY pos which guarantees us the order in which the transactions are loaded into the wallet.

The eventual goal is to use SQLite as a relational database with tables that store our keys and other metadata that can reference each other so that the wallet doesn’t just use the database as a storage mechanism. But that is still a work in progress.

This PR makes use of the last client opened features introduced in #32895 to determine when the old record style needs to be upgraded to the new transactions table. This should give us sufficient upgrade-downgrade-upgrade handling to not lose any funds.

Depends on #32763, #32895, #32985, #33032, #33033

DrahtBot added the label Wallet on Jul 21, 2025

DrahtBot commented at 10:52 pm on July 21, 2025: contributor

The following sections might be updated with supplementary metadata relevant to reviewers and maintainers.

Code Coverage & Benchmarks

For details see: https://corecheck.dev/bitcoin/bitcoin/pulls/33034.

Reviews

See the guideline for information on the review process.

Type	Reviewers
Concept ACK	w0xlt, rkrux

If your review is incorrectly listed, please copy-paste <!–meta-tag:bot-skip–> into the comment that the bot should ignore.

Conflicts

Reviewers, this pull request conflicts with the following ones:

#33033 (wallet, sqlite: Encapsulate SQLite statements in a RAII class by achow101)
#33032 (wallet, test: Replace MockableDatabase with in-memory SQLiteDatabase by achow101)
#32985 (wallet: Always rewrite tx records during migration by achow101)
#32895 (wallet: Prepare for future upgrades by recording versions of last client to open and decrypt by achow101)
#32685 (wallet: Allow read-only database access for info and dump commands by PeterWrighten)
#32636 (Split CWallet::Create() into CreateNew and LoadExisting by davidgumberg)
#32489 (wallet: Add exportwatchonlywallet RPC to export a watchonly version of a wallet by achow101)
#27865 (wallet: Track no-longer-spendable TXOs separately by achow101)
#19461 (multiprocess: Add bitcoin-gui -ipcconnect option by ryanofsky)
#19460 (multiprocess: Add bitcoin-wallet -ipcconnect option by ryanofsky)
#10102 (Multiprocess bitcoin by ryanofsky)

If you consider this pull request important, please also help to review the conflicting pull requests. Ideally, start with the one that should be merged first.

LLM Linter (✨ experimental)

Possible places where named args for integral literals may be used (e.g. func(x, /*named_arg=*/0) in C++, and func(x, named_arg=0) in Python):

sqlite3_prepare_v2(&m_db, stmt_text.c_str(), -1, &m_stmt, nullptr) in src/wallet/sqlite.cpp
SQLiteDatabase(fs::PathFromString(“mock/”), fs::PathFromString(“mock/wallet.dat”), DatabaseOptions(), SQLITE_OPEN_MEMORY) in src/wallet/test/util.cpp

2026-01-19

DrahtBot added the label CI failed on Jul 22, 2025

DrahtBot commented at 0:40 am on July 22, 2025: contributor

🚧 At least one of the CI tasks failed. Task lint: https://github.com/bitcoin/bitcoin/runs/46429203587 LLM reason (✨ experimental): The CI failure is caused by Python lint errors due to unresolved ‘self’ references in test files.

Try to run the tests locally, according to the documentation. However, a CI failure may still happen due to a number of reasons, for example:

Possibly due to a silent merge conflict (the changes in this pull request being incompatible with the current code in the target branch). If so, make sure to rebase on the latest commit of the target branch.
A sanitizer issue, which can only be found by compiling with the sanitizer and running the affected test.
An intermittent issue.

Leave a comment here, if you need help tracking down a confusing failure.

achow101 force-pushed on Jul 22, 2025

DrahtBot removed the label CI failed on Jul 22, 2025

achow101 force-pushed on Jul 23, 2025

w0xlt commented at 9:15 pm on July 23, 2025: contributor

Concept ACK

achow101 force-pushed on Jul 24, 2025

DrahtBot added the label Needs rebase on Jul 29, 2025

achow101 force-pushed on Jul 29, 2025

DrahtBot removed the label Needs rebase on Jul 29, 2025

DrahtBot added the label Needs rebase on Aug 8, 2025

achow101 force-pushed on Aug 8, 2025

DrahtBot removed the label Needs rebase on Aug 8, 2025

DrahtBot added the label Needs rebase on Aug 13, 2025

achow101 force-pushed on Aug 14, 2025

DrahtBot removed the label Needs rebase on Aug 14, 2025

DrahtBot added the label Needs rebase on Aug 16, 2025

achow101 force-pushed on Aug 16, 2025

achow101 force-pushed on Aug 19, 2025

DrahtBot removed the label Needs rebase on Aug 19, 2025

DrahtBot added the label Needs rebase on Sep 23, 2025

achow101 force-pushed on Sep 23, 2025

DrahtBot removed the label Needs rebase on Sep 24, 2025

rkrux commented at 8:14 am on October 9, 2025: contributor

Concept ACK a25557a

I have gone through the PR description and have skimmed over the commit messages as of now. At the outset, I find myself agreeing with the intent here. I am not aware of a strong enough reason to keep using SQLite as a KVS for all purposes, using its relational database properties for wallet transactions seem like a good start. I did notice some serialisation specifics of transaction properties in a couple previous PRs that I didn’t fully understand why were required; it could be a lot cleaner and explicit in implementation if using SQLite as an RDBMS can avoid the need for doing those specific serialisations.

DrahtBot added the label Needs rebase on Dec 2, 2025

achow101 force-pushed on Dec 10, 2025

DrahtBot removed the label Needs rebase on Dec 10, 2025

DrahtBot added the label CI failed on Dec 10, 2025

DrahtBot commented at 8:15 pm on December 10, 2025: contributor

🚧 At least one of the CI tasks failed. Task test max 6 ancestor commits: https://github.com/bitcoin/bitcoin/actions/runs/20110042688/job/57704599161 LLM reason (✨ experimental): CMake build failed with exit code 2 during project compilation (gmake reported an error).

Try to run the tests locally, according to the documentation. However, a CI failure may still happen due to a number of reasons, for example:

Possibly due to a silent merge conflict (the changes in this pull request being incompatible with the current code in the target branch). If so, make sure to rebase on the latest commit of the target branch.
A sanitizer issue, which can only be found by compiling with the sanitizer and running the affected test.
An intermittent issue.

Leave a comment here, if you need help tracking down a confusing failure.

achow101 force-pushed on Dec 10, 2025

DrahtBot removed the label CI failed on Dec 11, 2025

DrahtBot added the label Needs rebase on Dec 17, 2025

achow101 force-pushed on Dec 22, 2025

DrahtBot removed the label Needs rebase on Dec 22, 2025

DrahtBot added the label CI failed on Dec 23, 2025

achow101 force-pushed on Jan 3, 2026

DrahtBot removed the label CI failed on Jan 3, 2026

DrahtBot added the label Needs rebase on Jan 19, 2026

wallet: Pass replaces_txid to CommitTransaction outside of mapValue

Instead of updating mapValue with the "replaces_txid" value and passing
the updated mapValue to CommitTransaction, pass the replaces_txid
directly to CommitTransaction which updates mapValue as necessary.

This is prepration for removing mapValue.

098e4851bc

wallet: Pass comment and comment_to to CommitTransaction

Instead of passing these by setting them in mapValue, pass them directly
to CommitTransaction.

This is preparation for removing mapValue.

438856910b

wallet: Drop mapValue from CommitTransaction

The values previously passed in mapValue are now parameters to
CommitTransaction so there is no need for mapValue to be passed.

0a4702f55f

wallet: Make CWalletTx "from" and "message" member variables

Instead of storing "from" and "message" inside of mapValue, store these
explicitly as members of CWalletTx.

810288f12c

wallet: Make CWalletTx "comment" and "to" member variables

Instead of storing "comment" and "to" inside of mapValue, store these
expliclty as members of CWalletTx.

027e704d7b

wallet: Make CWalletTx "replaces_txid" and "replaced_by_txid" member variables

Instead of storing "replaces_txid" and "replaced_by_txid" as strings inside of
mapValue, store these expliclty as members of CWalletTx.

9710b9b2a2

wallet: Drop mapValue from CWalletTx

It doesn't make sense to be storing relevant metadata variables inside
of a string map in CWalletTx. All of the fields have been pulled out
into separate members, so there is no need for mapValue to stick around.

63a3b150b6

wallet: Drop vOrderForm from CommitTransaction

This parameter is only used to pass in the "Messages" from the GUI.
Instead of making it opaque by putting those into vOrderForm, use a
specific dedicated parameter for providing the messages.

556f09fd8d

wallet: Replace CWalletTx's vOrderForm with specific fields

vOrderForm contained 2 kinds of strings: BIP 21 messages, and BIP 70
Payment Requests. Instead of having both inside of a single vOrderForm
field that is opaque, split them into separate std::vector<std::string>
to contain this metadata.

15768d0e0d

achow101 force-pushed on Jan 19, 2026

walletdb: Decouple last client record from CLIENT_VERSION

Instead of tying the last client record with the node version, introduce
a separate wallet client version that will be increased as necessary
when new automatic upgrades are introduced.

58ce0e96a1

wallet: Introduce LastClientFeatures flags and LAST_OPENED_FEATURES record

LastClientFeatures are feature flags indicating what automatic upgrade
features supported by the last client that opened the wallet.

The LAST_OPENED_FEATURES record stores these flags and must be
set to match the exact features supported by the client that opens a
wallet.

610fa273d4

wallet: Record the supported features of the last client to decrypt a wallet 22337a39cf

wallet: Set last opened and decrypted features during migration

Set these flags to avoid the automatic upgrade after migrating.

54cfacbefd

wallet: Always rewrite tx records during migration

Since loading a wallet may change some parts of tx records (e.g. adding
nOrderPos), we should rewrite the records instead of copying them so
that the automatic upgrade does not need to be performed again when the
wallet is loaded.

7d739fcb57

bench, wallet: Make WalletMigration's setup WalletBatch scoped

WalletBatch needs to be in a scope so that it is destroyed before the
database is closed during migration.

c00e12ddef

test: Make duplicating MockableDatabases use cursor and batch

Instead of directly copying the stored records map when duplicating a
MockableDatabase, use a Cursor to read the records, and a Batch to write
them into the new database. This prepares for using SQLite as the
database backend for MockableDatabase.

4a988ed098

DrahtBot added the label CI failed on Jan 19, 2026

DrahtBot commented at 7:49 pm on January 19, 2026: contributor

🚧 At least one of the CI tasks failed. Task test max 6 ancestor commits: https://github.com/bitcoin/bitcoin/actions/runs/21148471093/job/60819287525 LLM reason (✨ experimental): Build failed: the CI cmake/make run exited with non-zero status (gmake: all target) after linking bitcoinkernel, indicating a generic build error.

Try to run the tests locally, according to the documentation. However, a CI failure may still happen due to a number of reasons, for example:

Possibly due to a silent merge conflict (the changes in this pull request being incompatible with the current code in the target branch). If so, make sure to rebase on the latest commit of the target branch.
A sanitizer issue, which can only be found by compiling with the sanitizer and running the affected test.
An intermittent issue.

Leave a comment here, if you need help tracking down a confusing failure.

DrahtBot removed the label Needs rebase on Jan 19, 2026

wallet, bench: Use TestingSetup in CoinSelection benchmark db8d8bab7f

wallet: Make Mockable{Database,Batch} subclasses of SQLite classes

The mocking functionality of MockableDatabase, MockableBatch, and
MockableCursor was not really being used. These are changed to be
subclasses of their respective SQLite* classes and will use in-memory
SQLite databases so that the tests are more representative of actual
database behavior.

MockableCursor is removed as there are no overrides needed in
SQLiteCursor for the tests.

38a5670b2d

walletdb: Remove m_mock from SQLiteDatabase ee2b763d92

sqlite: Add SQLiteStatement RAII class

This class will be used to encapsulate a sqlite3_stmt

bfba6ba187

sqlite: Make Column template function 4f7c372645

sqlite: Refactor ReadPragmaInteger to use SQLiteStatement da8a9af870

sqlite: Use SQLiteStatement in PRAGMA integrity_check b099a62572

sqlite: Use SQLiteStatement in check_main_stmt eea5062d5d

sqlite: Have SQLiteCursor store SQLiteStatement fba50ea6ad

sqlite: Replace remaining sqlite3_stmt usage with SQLiteStatement 337752f0ac

sqlite: Refactor common writing code from WriteKey and ExecStatement

WriteKey and ExecStatement use the same code for the actual execution of
the statement. This is refactored into a separate function, also called
ExecStatement, and the original ExecStatement renamed to
ExecEraseStatement as it is only used by the erase functions.

dc2f580cb6

sqlite: Inline BindBlobToStatement and SpanFromBlob 8b94ab73f1

util: Add span constructor to (W)Txid c7de7a7b5c

sqlite: Make SQLiteStatement::Bind a template function

We will need to bind data types other than blob

1c6ce965de

sqlite: Make SQLiteDatabase::Column an std::optional

To handle columns containing NULL values, Column needs to return some
value representing NULL, so make it a std::optional.

24798ff5f0

sqlite: Add additional blob types to SQLiteStatement::Column

Not all blob data types fit in ColumnBlob, so we need additional
template type requirements to match those.

3ea04e2ea4

wallet: Add Un/Serialize to TxState structs 873f794ab3

wallet: Add new serialization format functions for TxState 9cbaef5304

sqlite: Create a 'transactions' table if it does not exist 0ca6949f46

sqlite: Add functions and statements for transactions table 1e53a2cb43

sqlite: Iterate the transactions table with SQLiteCursor 8c291a5d36

walletdb: Add functions to modify transactions table b8705ca6f0

test: Include sql transactions table in DuplicateMockDatabase e7c9e23f5f

wallet: Also write to the new transactions table d3978d6d31

wallet: Perform automatic upgrade to using the transactions table 35858421d8

walletdb: Load from transactions table and use original tx for upgrade

When loading a wallet, always load from the transactions table, except
when loading a legacy wallet for migration.

The original 'tx' records are only used to upgrade to the transactions
table if an upgraded is necessary.

4923f6b927

achow101 force-pushed on Jan 19, 2026

wallet: Store transactions in a separate sqlite table #33034

Code Coverage & Benchmarks

Reviews

Conflicts

LLM Linter (✨ experimental)