Bump rocksdb to version 8.6.7

[aleksraiden]

A new release, Rocksdb 8.6.7. This is an bug-fix release. A full changelog - https://github.com/facebook/rocksdb/releases/tag/v8.6.7

Bug-fix

• Fix bug with an error if user have a large number of levels
• Add a fix for async_io where during seek, when reading a block for seeking a target key in a file without any readahead, the iterator aligned the read on a page boundary and reading more than necessary. This increased the storage read bandwidth usage
• Fix a bug where iterator may return incorrect result for DeleteRange
• Fix a bug where if there is an error reading from offset 0 of a file from L1+ and that the file is not the first file in the sorted run
• Fixed a bug where rocksdb.file.read.verify.file.checksums.micros is not populated
• Fix a bug with atomic_flush=true that can cause DB
• Fix a bug where RocksDB (with atomic_flush=false) can delete output SST files of pending flushes when a previous concurrent flush fails
• Fix bug with compressed secondary cache
• Fixed a bug where compaction read under non direct IO still falls back to RocksDB internal prefetching

Behavior Changes

• For Universal Compaction users, periodic compaction (option periodic_compaction_seconds) will be set to 30 days by default if block based table is used
• Mark Options::access_hint_on_compaction_start related APIs as deprecated
• Add a column family option default_temperature that is used for file reading accounting purpose, such as io statistics, for files that don't have an explicitly set temperature.

New

• compaction_readahead_size 's default value is changed from 0 to 2MB.
• Added enhanced data integrity checking on SST files
• Add support to allow enabling / disabling user-defined timestamps feature for an existing column family in combination with the in-Memtable only feature.
• Add a new compression option CompressionOptions::checksum for enabling ZSTD's checksum feature to detect corruption during decompression

[mapleFU]

Description below is incomplete:

Fix a bug with atomic_flush=true that can cause DB to stuck after a flush fails (https://github.com/facebook/rocksdb/pull/11872).

Rest LGTM

[aleksraiden]

Description below is incomplete:

Fix a bug with atomic_flush=true that can cause DB to stuck after a flush fails (https://github.com/facebook/rocksdb/pull/11872).

Rest LGTM

You are right, my mistake, I try to shorter a desc for PR.

[mapleFU]

General looks good to me. I'm on the way, I'll review a round carefully when I get home tonight :-)

[mapleFU]

Statistics rocksdb.sst.read.micros now includes time spent on multi read and async read into the file

Since this can help our statistics if we use this. (But seems that this is unused)

Other changes in rocksdb changelog seems doesn't have any risks.

[aleksraiden]

Statistics rocksdb.sst.read.micros now includes time spent on multi read and async read into the file

Since this can help our statistics if we use this. (But seems that this is unused)

In all latest release, as I see, a rocksdb team has an intensive work about extending stats and added a lot of new metrics.