From e440ac7ef3b6f3ad1cd8fc7027cece40413202d9 Mon Sep 17 00:00:00 2001 From: Gregory Maxwell Date: Fri, 6 Jan 2017 11:49:59 +0000 Subject: Introduce assumevalid setting to skip presumed valid scripts. This disentangles the script validation skipping from checkpoints. A new option is introduced "assumevalid" which specifies a block whos ancestors we assume all have valid scriptsigs and so we do not check them when they are also burried under the best header by two weeks worth of work. Unlike checkpoints this has no influence on consensus unless you set it to a block with an invalid history. Because of this it can be easily be updated without risk of influencing the network consensus. This results in a massive IBD speedup. This approach was independently recommended by Peter Todd and Luke-Jr since POW based signature skipping (see PR#9180) does not have the verifiable properties of a specific hash and may create bad incentives. The downside is that, like checkpoints, the defaults bitrot and older releases will sync slower. On the plus side users can provide their own value here, and if they set it to something crazy all that will happen is more time will be spend validating signatures. Checkblocks and checklevel are also moved to the hidden debug options: Especially now that checkblocks has a low default there is little need to change these settings, and users frequently misunderstand them as influencing security or IBD speed. By hiding them we offset the space added by this new option. --- doc/release-notes.md | 23 ++++++++++++++++++++++- doc/release-process.md | 5 +++++ 2 files changed, 27 insertions(+), 1 deletion(-) (limited to 'doc') diff --git a/doc/release-notes.md b/doc/release-notes.md index 7d0d689684..cb3f7934ce 100644 --- a/doc/release-notes.md +++ b/doc/release-notes.md @@ -49,7 +49,7 @@ Low-level RPC changes than two arguments. Removal of Priority Estimation ------------------------------- +------------------------------- - Estimation of "priority" needed for a transaction to be included within a target number of blocks has been removed. The rpc calls are deprecated and will either @@ -71,6 +71,27 @@ P2P connection management - New connections to manually added peers are much faster. +Introduction of assumed-valid blocks +------------------------------------- + +- A significant portion of the initial block download time is spent verifying + scripts/signatures. Although the verification must pass to ensure the security + of the system, no other result from this verification is needed: If the node + knew the history of a given block were valid it could skip checking scripts + for its ancestors. + +- A new configuration option 'assumevalid' is provided to express this knowledge + to the software. Unlike the 'checkpoints' in the past this setting does not + force the use of a particular chain: chains that are consistent with it are + processed quicker, but other chains are still accepted if they'd otherwise + be chosen as best. Also unlike 'checkpoints' the user can configure which + block history is assumed true, this means that even outdated software can + sync more quickly if the setting is updated by the user. + +- Because the validity of a chain history is a simple objective fact it is much + easier to review this setting. As a result the software ships with a default + value adjusted to match the current chain shortly before release. The use + of this default value can be disabled by setting -assumevalid=0 0.14.0 Change log ================= diff --git a/doc/release-process.md b/doc/release-process.md index 61f05b0771..399ed25c91 100644 --- a/doc/release-process.md +++ b/doc/release-process.md @@ -13,6 +13,11 @@ Before every minor and major release: * Update version in sources (see below) * Write release notes (see below) * Update `src/chainparams.cpp` nMinimumChainWork with information from the getblockchaininfo rpc. +* Update `src/chainparams.cpp` defaultAssumeValid with information from the getblockhash rpc. + - The selected value must not be orphaned so it may be useful to set the value two blocks back from the tip. + - Testnet should be set some tens of thousands back from the tip due to reorgs there. + - This update should be reviewed with a reindex-chainstate with assumevalid=0 to catch any defect + that causes rejection of blocks in the past history. Before every major release: -- cgit v1.2.3