Commit Graph

113 Commits

Author SHA1 Message Date
Howard Chu d6f5d543a5 Fix Issue #855
Use the same size dirty list for both 64 and 32 bit.
2016-06-07 23:38:47 +01:00
Howard Chu 66b1e13aa7 mdb_drop optimization
If we know there are no sub-DBs and no overflow pages, skip leaf scan.
2016-04-09 20:44:05 +01:00
Howard Chu d7ea7d9a23 Merge branch 'performance' into master 2016-04-05 21:13:16 +01:00
Howard Chu 6225716f3c More outputs consolidation
Also bumped DB VERSION to 1
Another significant speedup and space savings:
Get rid of global_output_indices, remove indirection from output to keys

This is the change warptangent described on irc but never got to finish.
2016-04-05 20:55:12 +01:00
Riccardo Spagni 40974b155e
fix building on FreeBSD 2016-03-21 19:37:07 +02:00
Riccardo Spagni 1800d611a2
bump miniupnpc API version number 2016-03-21 15:05:50 +02:00
Riccardo Spagni a4242c42b2
update miniupnpc 2016-03-21 13:37:54 +02:00
Howard Chu 8860b7497f MDB_VL32 - increase max write txn size 2016-02-17 20:41:34 +00:00
Howard Chu 31e4e8c3a3 Resync with master 2016-02-16 23:38:04 +00:00
warptangent 9832d18dca
cmake: Include OpenSSL libraries in static linking 2016-02-12 16:54:14 -08:00
warptangent 7205210b0f
cmake: Fix unbound config compile settings
This allows the OpenSSL function checks to compile in unbound's CMake
configuration.

Otherwise, the functions SHA256() and EVP_sha512() won't be called from
libunbound as possible algorithms.

They had not been compiling because static OpenSSL libraries were being
used, along with lack of -ldl. The static library preference is
unnecessary for the checks, so use default suffixes ordering for
CMAKE_FIND_LIBRARY_SUFFIXES when building unbound.

Related files:
configure_checks.cmake
external/unbound/validator/val_secalgo.c
  secalgo_ds_digest(), setup_key_digest()
2016-02-12 15:02:06 -08:00
Howard Chu f98fe467e9 MDB_VL32 change overflow page scan
Just check the requested page, don't worry about any other pages
2016-01-28 13:59:02 +00:00
Howard Chu 7e65cc0aba MDB_VL32 Fix off-by-one in mdb_midl_shrink 2016-01-28 04:15:45 +00:00
Howard Chu 222e30ec97 MDB_VL32 Fix another 32bit overflow 2016-01-27 14:14:27 +00:00
Howard Chu 3ca6e853f6 Tweak mdb_strerror msg buffer 2016-01-27 14:14:11 +00:00
Howard Chu 78ad970a2c MDB_VL32 Fix d2a5f72f73
VirtualAlloc is not for MDB_VL32
2016-01-27 14:14:02 +00:00
Howard Chu 1cff3974d8 WIN64 needs off_t redefined too 2016-01-20 01:26:26 +00:00
Howard Chu d2a5f72f73 Fix --db-sync-mode on Windows64
only "fastest" mode was working, others would SEGV.
2016-01-16 16:11:11 +00:00
Howard Chu 0b6be4114d There has never been any such function "fnctl"
It was only a typo of fcntl. Get rid of this test.
2016-01-12 21:19:15 +00:00
Howard Chu fcf31f3ff3 Cleanup symbol clashes, stdlib header 2016-01-11 23:09:50 +00:00
Howard Chu 28b82718ff Avoid cmake-3.x block comment syntax 2016-01-03 08:22:06 +00:00
Howard Chu 462658df4f Get rid of doubly-defined NDEBUG 2016-01-02 02:01:58 +00:00
Riccardo Spagni 88b0fb14c8
let miniupnp's cmakelist handle defaults 2015-12-31 15:15:57 +02:00
Riccardo Spagni 2c82eb71e7
fix miniupnpc cmakelist 2015-12-31 09:06:11 +02:00
Riccardo Spagni de03926850
updated copyright year 2015-12-31 08:39:56 +02:00
Howard Chu 10ae0032d8 isblank doesn't need a special case test 2015-12-31 05:34:29 +00:00
Howard Chu a090ee97f9 Fix 3edbf57b62
The test for isblank was being ignored
2015-12-31 05:31:42 +00:00
Riccardo Spagni 3edbf57b62
fix missing unbound tests 2015-12-30 13:20:06 +02:00
Riccardo Spagni 2d43ae8063
update unbound, fix unbound openssl issue on OS X 2015-12-30 12:57:50 +02:00
Riccardo Spagni 32a26332f8
no longer need to pass the size to rapidjson 2015-12-30 12:25:29 +02:00
Riccardo Spagni bd8e0fd2a1
add missing miniupnpc files, modify cmake to not build miniupnpc tests and to fix an issue with finding miniupnpcstrings 2015-12-30 10:00:50 +02:00
Riccardo Spagni 7da9905589
updated miniupnpc 2015-12-30 09:56:25 +02:00
Howard Chu b773e66877 MDB_VL32 - resync with master
WIN32 - close file mapping handle in env_close
cursor_unref - ignore cursor with empty stack
2015-12-28 20:46:48 +00:00
Howard Chu ba4e217c90 Update liblmdb, unify 32/64 sources 2015-12-25 08:56:03 +00:00
moneromooo-monero 234f576565
miniupnpc: quick fix for buffer overflow
http://talosintel.com/reports/TALOS-2015-0035/

reported by palexander on IRC
2015-10-09 15:20:05 +01:00
moneromooo-monero 5d0cb73356
unbound: do not try to link against libevent
It is now not needed anymore, as we use minievent instead
2015-10-05 20:57:52 +01:00
moneromooo-monero 3c10239327
unbound: use the mini event fallback implementation
Using libevent seems to have high peaks of file descriptor use,
which can cause failure to create fds in other parts of bitmonerod.
The fallback implementation seems to run fine in a significantly
tighter file descriptor limit.
2015-08-30 15:21:24 +01:00
Riccardo Spagni 9672ac0812 Revert "re-re-add Windows DNS bug fix, per c0de96f"
This reverts commit f4ba92bccc.
2015-08-18 02:26:02 +02:00
Riccardo Spagni f4ba92bccc
re-re-add Windows DNS bug fix, per c0de96f 2015-08-14 19:33:43 +02:00
Riccardo Spagni 93944333c5
update unbound 2015-08-14 19:12:19 +02:00
Riccardo Spagni 35b62fbd8e
hyc accidentally typo'd...we shall never speak of this again 2015-07-16 13:32:20 +02:00
Riccardo Spagni a4254a48f1
updated vl32 to current 2015-07-16 12:46:15 +02:00
Riccardo Spagni e6ab2dfdf3
updated in-source lmdb 2015-07-16 11:33:10 +02:00
Thomas Winget ea08c768e2 open() flag O_DSYNC isn't on BSD, use O_SYNC
If the detected OS is FreeBSD, tell LMDB to compile with
MDB_DSYNC=O_SYNC instead of the default O_DSYNC, as BSD does not
implement this flag.
2015-07-16 11:22:39 +02:00
NoodleDoodleNoodleDoodleNoodleDoodleNoo e5d2680094 ** CHANGES ARE EXPERIMENTAL (FOR TESTING ONLY)
Bockchain:
1. Optim: Multi-thread long-hash computation when encountering groups of blocks.
2. Optim: Cache verified txs and return result from cache instead of re-checking whenever possible.
3. Optim: Preload output-keys when encoutering groups of blocks. Sort by amount and global-index before bulk querying database and multi-thread when possible.
4. Optim: Disable double spend check on block verification, double spend is already detected when trying to add blocks.
5. Optim: Multi-thread signature computation whenever possible.
6. Patch: Disable locking (recursive mutex) on called functions from check_tx_inputs which causes slowdowns (only seems to happen on ubuntu/VMs??? Reason: TBD)
7. Optim: Removed looped full-tx hash computation when retrieving transactions from pool (???).
8. Optim: Cache difficulty/timestamps (735 blocks) for next-difficulty calculations so that only 2 db reads per new block is needed when a new block arrives (instead of 1470 reads).

Berkeley-DB:
1. Fix: 32-bit data errors causing wrong output global indices and failure to send blocks to peers (etc).
2. Fix: Unable to pop blocks on reorganize due to transaction errors.
3. Patch: Large number of transaction aborts when running multi-threaded bulk queries.
4. Patch: Insufficient locks error when running full sync.
5. Patch: Incorrect db stats when returning from an immediate exit from "pop block" operation.
6. Optim: Add bulk queries to get output global indices.
7. Optim: Modified output_keys table to store public_key+unlock_time+height for single transaction lookup (vs 3)
8. Optim: Used output_keys table retrieve public_keys instead of going through output_amounts->output_txs+output_indices->txs->output:public_key
9. Optim: Added thread-safe buffers used when multi-threading bulk queries.
10. Optim: Added support for nosync/write_nosync options for improved performance (*see --db-sync-mode option for details)
11. Mod: Added checkpoint thread and auto-remove-logs option.
12. *Now usable on 32-bit systems like RPI2.

LMDB:
1. Optim: Added custom comparison for 256-bit key tables (minor speed-up, TBD: get actual effect)
2. Optim: Modified output_keys table to store public_key+unlock_time+height for single transaction lookup (vs 3)
3. Optim: Used output_keys table retrieve public_keys instead of going through output_amounts->output_txs+output_indices->txs->output:public_key
4. Optim: Added support for sync/writemap options for improved performance (*see --db-sync-mode option for details)
5. Mod: Auto resize to +1GB instead of multiplier x1.5

ETC:
1. Minor optimizations for slow-hash for ARM (RPI2). Incomplete.
2. Fix: 32-bit saturation bug when computing next difficulty on large blocks.

[PENDING ISSUES]
1. Berkely db has a very slow "pop-block" operation. This is very noticeable on the RPI2 as it sometimes takes > 10 MINUTES to pop a block during reorganization.
   This does not happen very often however, most reorgs seem to take a few seconds but it possibly depends on the number of outputs present. TBD.
2. Berkeley db, possible bug "unable to allocate memory". TBD.

[NEW OPTIONS] (*Currently all enabled for testing purposes)
1. --fast-block-sync arg=[0:1] (default: 1)
	a. 0 = Compute long hash per block (may take a while depending on CPU)
	b. 1 = Skip long-hash and verify blocks based on embedded known good block hashes (faster, minimal CPU dependence)
2. --db-sync-mode arg=[[safe|fast|fastest]:[sync|async]:[nblocks_per_sync]] (default: fastest:async:1000)
	a. safe = fdatasync/fsync (or equivalent) per stored block. Very slow, but safest option to protect against power-out/crash conditions.
	b. fast/fastest = Enables asynchronous fdatasync/fsync (or equivalent). Useful for battery operated devices or STABLE systems with UPS and/or systems with battery backed write cache/solid state cache.
	Fast    - Write meta-data but defer data flush.
	Fastest - Defer meta-data and data flush.
	Sync    - Flush data after nblocks_per_sync and wait.
	Async   - Flush data after nblocks_per_sync but do not wait for the operation to finish.
3. --prep-blocks-threads arg=[n] (default: 4 or system max threads, whichever is lower)
        Max number of threads to use when computing long-hash in groups.
4. --show-time-stats arg=[0:1] (default: 1)
	Show benchmark related time stats.
5. --db-auto-remove-logs arg=[0:1] (default: 1)
	For berkeley-db only. Auto remove logs if enabled.

**Note: lmdb and berkeley-db have changes to the tables and are not compatible with official git head version.
	At the moment, you need a full resync to use this optimized version.

[PERFORMANCE COMPARISON]
**Some figures are approximations only.
Using a baseline machine of an i7-2600K+SSD+(with full pow computation):
1. The optimized lmdb/blockhain core can process blocks up to 585K for ~1.25 hours + download time, so it usually takes 2.5 hours to sync the full chain.
2. The current head with memory can process blocks up to 585K for ~4.2 hours + download time, so it usually takes 5.5 hours to sync the full chain.
3. The current head with lmdb can process blocks up to 585K for ~32 hours + download time and usually takes 36 hours to sync the full chain.

Averate procesing times (with full pow computation):
lmdb-optimized:
1. tx_ave = 2.5 ms / tx
2. block_ave = 5.87 ms / block
memory-official-repo:
1. tx_ave = 8.85 ms / tx
2. block_ave = 19.68 ms / block
lmdb-official-repo (0f4a036437)
1. tx_ave = 47.8 ms / tx
2. block_ave = 64.2 ms / block

**Note: The following data denotes processing times only (does not include p2p download time)
lmdb-optimized processing times (with full pow computation):
1. Desktop,  Quad-core / 8-threads 2600k  (8Mb) - 1.25 hours processing time (--db-sync-mode=fastest:async:1000).
2. Laptop,   Dual-core / 4-threads U4200  (3Mb) - 4.90 hours processing time (--db-sync-mode=fastest:async:1000).
3. Embedded, Quad-core / 4-threads Z3735F (2x1Mb) - 12.0 hours processing time (--db-sync-mode=fastest:async:1000).

lmdb-optimized processing times (with per-block-checkpoint)
1. Desktop,  Quad-core / 8-threads 2600k  (8Mb) - 10 minutes processing time (--db-sync-mode=fastest:async:1000).

berkeley-db optimized processing times (with full pow computation)
1. Desktop, Quad-core / 8-threads 2600k  (8Mb) - 1.8 hours processing time (--db-sync-mode=fastest:async:1000).
2. RPI2. Improved from estimated 3 months(???) into 2.5 days (*Need 2AMP supply + Clock:1Ghz + [usb+ssd] to achieve this speed) (--db-sync-mode=fastest:async:1000).

berkeley-db optimized processing times (with per-block-checkpoint)
1. RPI2. 12-15 hours (*Need 2AMP supply + Clock:1Ghz + [usb+ssd] to achieve this speed) (--db-sync-mode=fastest:async:1000).
2015-07-15 23:20:16 -07:00
Riccardo Spagni 0f4a036437
re-add Windows DNS bug fix, per c0de96f8bd 2015-06-02 11:33:04 +02:00
Riccardo Spagni 6a1190792b
update libunbound 2015-05-31 16:36:48 +02:00
Thomas Winget bed2331904
Require BerkeleyDB to be installed (for now) if building non-static 2015-04-22 01:51:14 -04:00
Riccardo Spagni 41f0a8fe4d
update lmdb64 2015-04-14 21:07:39 +02:00
Thomas Winget a8bc7182ea
Merge BlockchainDB into upstream 2015-04-07 17:56:18 -04:00