⚠️ otech47 opened an issue: "Running out of memory on a 2GB box - Initializing chainstate Chainstate [ibd] @ height -1 (null)"
(https://github.com/bitcoin/bitcoin/issues/31573)
I have a server running a pruned bitcoind node as a Docker container with 2G RAM. There are some other services using maybe 500MB of the RAM but I got into a state where bitcoind tries to initialize the chainstate but then starts sucking up all the available RAM so it gets killed with OOM error. Docker then restarts it and it loops like this forever

I tried to configure bitcoind to be constrained to low memory but it doesn't seem to have any effect

```
[main]
printtoconsole=1
rpcallowip
...

🤔 furszy reviewed a pull request: "wallet: refactor: various master key encryption cleanups"
(https://github.com/bitcoin/bitcoin/pull/31398#pullrequestreview-2524287853)
Concept ACK, nice code dedup.

💬 danielabrozzoni commented on pull request "rpc: combinerawtransaction now rejects unmergeable transactions":
(https://github.com/bitcoin/bitcoin/pull/31298#discussion_r1898656359)
I don't think it's needed to clear the script witness, as it's not taken into account when calculating the txid

📝 SecondPort opened a pull request: "Fix memory issue during chainstate initialization on 2GB box"
(https://github.com/bitcoin/bitcoin/pull/31574)
Fixes #31573

---

For more details, open the [Copilot Workspace session](https://copilot-workspace.githubnext.com/bitcoin/bitcoin/pull/31574?shareId=adfd9f84-4320-4707-98a2-35d355535cbf).

✅ SecondPort closed a pull request: "Fix memory issue during chainstate initialization on 2GB box"
(https://github.com/bitcoin/bitcoin/pull/31574)

📝 SecondPort reopened a pull request: "Fix memory issue during chainstate initialization on 2GB box"
(https://github.com/bitcoin/bitcoin/pull/31574)
Fixes #31573

✅ SecondPort closed a pull request: "Fix memory issue during chainstate initialization on 2GB box"
(https://github.com/bitcoin/bitcoin/pull/31574)

📝 SecondPort opened a pull request: "Fix memory issue during chainstate initialization on 2GB box"
(https://github.com/bitcoin/bitcoin/pull/31575)
Fixes #31573

Reduce memory usage during chainstate initialization on low-memory systems.

* **src/txdb.h**
- Reduce the default `dbcache` value to 100 MiB.
* **src/validation.h**
- Add a function to adjust `dbcache` to 100 MiB if the system has less than 4GB of RAM.
* **doc/reduce-memory.md**
- Update the documentation to reflect the new default `dbcache` value.
- Add information about the memory check for systems with less than 4GB of RAM.

💬 l0rinc commented on pull request "Fix memory issue during chainstate initialization on 2GB box":
(https://github.com/bitcoin/bitcoin/pull/31574#discussion_r1898673122)
NACK

💬 l0rinc commented on pull request "Fix memory issue during chainstate initialization on 2GB box":
(https://github.com/bitcoin/bitcoin/pull/31575#issuecomment-2563957123)
> GetTotalSystemMemory’ was not declared in this scope

Please compile it locally before pushing.

📝 hebasto opened a pull request: "test: Move `script_assets_tests` into its own suite"
(https://github.com/bitcoin/bitcoin/pull/31576)
This PR ensures that the `script_assets_tests` test case is explicitly reported as "Skipped" when it is not run, making it clearer when running the test suite with `ctest`:

- on the master branch @ 9355578a77978a0c2f189bd7315a2883142d8119:
```
$ env -u DIR_UNIT_TEST_DATA ctest --test-dir build -j 16 -R "^script_"
Internal ctest changing into directory: /home/hebasto/git/bitcoin/build
Test project /home/hebasto/git/bitcoin/build
Start 87: script_tests
Start 83: script_p2sh_tests
...

📝 hebasto converted_to_draft a pull request: "test: Move `script_assets_tests` into its own suite"
(https://github.com/bitcoin/bitcoin/pull/31576)
This PR ensures that the `script_assets_tests` test case is explicitly reported as "Skipped" when it is not run, making it clearer when running the test suite with `ctest`:

- on the master branch @ 9355578a77978a0c2f189bd7315a2883142d8119:
```
$ env -u DIR_UNIT_TEST_DATA ctest --test-dir build -j 16 -R "^script_"
Internal ctest changing into directory: /home/hebasto/git/bitcoin/build
Test project /home/hebasto/git/bitcoin/build
Start 87: script_tests
Start 83: script_p2sh_tests
...

💬 mzumsande commented on pull request "validation: write chainstate to disk every hour":
(https://github.com/bitcoin/bitcoin/pull/30611#discussion_r1898689788)
This comment could be adjusted - `fPeriodicWrite` is no longer just about the block index.

👋 hebasto's pull request is ready for review: "test: Move `script_assets_tests` into its own suite"
(https://github.com/bitcoin/bitcoin/pull/31576)

💬 deepgrewal commented on pull request "qa: Limit `-maxconnections` in tests":
(https://github.com/bitcoin/bitcoin/pull/31537#issuecomment-2563996883)
I do not agree with this change and it should not be merged.

💬 davidgumberg commented on issue "b-msghand invoked oom-killer: Master (v28.99) crashing during IBD":
(https://github.com/bitcoin/bitcoin/issues/31561#issuecomment-2564008796)
> The node in question finished sync after I ran it inside heaptrack and is humming along just fine now. I printed out the flamegraph although I have trouble interpreting it:
>
> [ flamegraph.svg]

At first glance it seems pretty unusual that so much time (~40% of samples) is spent in the call stack above `CScriptCheck::operator()` seemingly doing allocations.

💬 mzumsande commented on pull request "qa: Limit `-maxconnections` in tests":
(https://github.com/bitcoin/bitcoin/pull/31537#issuecomment-2564016601)
> I do not agree with this change and it should not be merged.

"[A NACK needs to include a rationale why the change is not worthwhile. NACKs without accompanying reasoning may be disregarded.](https://github.com/bitcoin/bitcoin/blob/master/CONTRIBUTING.md)"

💬 l0rinc commented on pull request "RPC: Add reserve member function to `UniValue` and use it in `getblock` RPC":
(https://github.com/bitcoin/bitcoin/pull/31179#issuecomment-2564016916)
@ismaelsadeeq, I ran your benchmarks from https://github.com/bitcoin/bitcoin/pull/31179#issuecomment-2446067820, but I only see a 7% speedup in the `TxVerbosity::SHOW_TXID` bench - while the rest were exactly the same as before.
But if it's *that* important here, we could easily get a 30% speedup as well there by deduplicating the `GetSerializeSize` calls:
```C++
static int64_t GetBlockWeight(const size_t size_no_witness, const size_t size_with_witness)
{
return size_no_witness * (WITNE
...

🤔 furszy reviewed a pull request: "validation: Send correct notification during snapshot completion"
(https://github.com/bitcoin/bitcoin/pull/31556#pullrequestreview-2524378694)
Code review ACK bc43ecaf6dc

💬 brunoerg commented on pull request "test: descriptor: fix test for `MaxSatisfactionWeight`":
(https://github.com/bitcoin/bitcoin/pull/31570#discussion_r1898710992)
Done.

💬 brunoerg commented on pull request "test: descriptor: fix test for `MaxSatisfactionWeight`":
(https://github.com/bitcoin/bitcoin/pull/31570#issuecomment-2564026547)
Force-pushed addressing https://github.com/bitcoin/bitcoin/pull/31570#discussion_r1898539802