Sears Russell
dc1f799a37
TarrayListAlloc() and TarrayListExtend() now generate O(1) bytes of log entries; they use the new multipage operation type.
2010-05-11 23:34:13 +00:00
Sears Russell
820f706faf
TarrayListExtend now zeros out the contents of the newly allocated records (this is implemented by having fixed.c zero out the records)
2010-05-11 23:27:22 +00:00
Sears Russell
2d645beb42
cleaned up multipage operations a bit; added stasis_segments_enabled flag
2010-05-11 22:01:37 +00:00
Sears Russell
07df7c064a
add support for multi-page operations
2010-05-11 00:17:07 +00:00
Sears Russell
31d045efba
add link to my dissertation
2010-05-05 19:27:27 +00:00
Sears Russell
85ad4dd316
updated getting started text
2010-05-05 18:54:45 +00:00
Sears Russell
ffa4f18283
add tryToWriteBackPage. concurrentBufferManager is now compatible with FORCE mode transactions.
2010-04-28 00:05:00 +00:00
Sears Russell
c7fa929e71
This fixes in Stasis issue 22.
...
Replacement policies now track the pin count of each page that they interact with. This behavior used to live in bufferHash, and had never been ported to concurrentBufferManager.
For various reasons, the pin counts have to be updated atomically with the LRU state. This would be very difficult to do inside of concurrentBufferManager, which sometimes needs to add and remove pages to replacement policies without holding any other latches.
2010-04-20 01:09:32 +00:00
Sears Russell
887832582c
expose stasis_handle_t up through the buffer manager; add sequential I/O hinting; fix stupid busy-wait in prefetch
2010-04-12 20:54:10 +00:00
Sears Russell
7c5eb2a7e9
added dup method to io handles
2010-04-08 23:50:01 +00:00
Sears Russell
ff2f2e9e1e
add TregionPrefetch()
2010-03-31 22:41:33 +00:00
Sears Russell
3c27ed566e
add prefetch primitives
2010-03-29 22:29:30 +00:00
Sears Russell
8a5bbd9a38
silence compiler warning
2010-03-29 22:28:32 +00:00
Sears Russell
6deb34f5b2
fix handling of loadPageForOperation during recovery; initialize rid.size in slottedLast
2010-03-22 19:56:55 +00:00
Sears Russell
5ffd1110e9
Added concept of "transaction fingerprints"
...
The region allocator marks freed regions "Condemned" before actually freeing them. Condemned regions are associated
with a transaction; before they can be reused / coalesced the transaction that condemned them must complete.
The region allocator used "TisActiveTransaction(xid)" to determine whether a given transaction had completed. The
problem is that Stasis now reuses transaction id's, and condemned regions were never reclaimed in practice. A transaction
fingerprint is a unique identifier (currently transaction id, rec_lsn) that can be used to distinguish transactions that
did not execute concurrently.
I considered an alternative design based upon end-of-transaction actions. This would have worked as well, but would
have been a more invasive change.
2010-02-18 19:12:46 +00:00
Sears Russell
4e100479d1
Add TarrayListDealloc() to arrayList. Remove one region allocation from each array list alloc. Add unit test for dealloc.
...
Also, removed some old compensations cruft.
2010-02-18 04:12:39 +00:00
Sears Russell
98e0cb1fab
implement TregionSize(), add TregionForce()
2010-02-12 16:44:28 +00:00
Sears Russell
0afb9177ac
C++ compatibility fix
2010-02-11 22:18:49 +00:00
Sears Russell
1ff9afb753
added new constant: INVALID_SIZE
2010-02-11 20:36:33 +00:00
Sears Russell
49979ffa52
unit test for libredblack
2010-01-21 23:48:43 +00:00
Sears Russell
dbf253275f
fix nested top action log race; because of a refactoring mistake, it was releasing a mutex without acquiring it...
2010-01-21 20:27:35 +00:00
Sears Russell
250223086c
reordered includes for compatibility with gcc3
2010-01-21 00:03:17 +00:00
Sears Russell
f1f1d888e4
avoid use of -Wno-variadic-macros, which causes an error with gcc 3.
2010-01-20 23:51:54 +00:00
Sears Russell
b1f7d1947a
more log refactoring. safeWrites now holds a latch across reserve_entry, write_entry, and write_entry_done
2010-01-19 02:14:09 +00:00
Sears Russell
c8cc188b71
unsigned int -> uint32_t
2010-01-19 01:47:13 +00:00
Sears Russell
d71e248291
removed freeLogEntry
2010-01-11 20:41:59 +00:00
Sears Russell
c30aaa3a1a
fix compiler warning; -1 is not comparable to entry->type
2010-01-07 21:50:17 +00:00
Sears Russell
1af589dd18
remove malloc and memcpy from the in_memory_log read path
2009-12-31 23:25:56 +00:00
Sears Russell
1409a9eb27
more refactoring; log handles now manage their log entries memory
2009-12-31 23:01:37 +00:00
Sears Russell
a5788688c8
refactor getUpdateArgs; remove a bunch of casts, fix a bunch of warnings
2009-12-31 20:20:47 +00:00
Sears Russell
9aa1d9d652
pull memcpy out of allocUpdateEntry()
2009-12-31 19:30:40 +00:00
Sears Russell
32e40c949d
added log modes to sequentialThroughput benchmark, added methods for concurrent appends to log api (and stubbed them out)
2009-12-31 19:02:34 +00:00
Sears Russell
7ff80f37ef
add deferred, end of transaction callbacks. addresses issue 15
2009-12-15 06:30:37 +00:00
Sears Russell
efc6a0eabd
add missing function declaration
2009-12-10 01:53:15 +00:00
Sears Russell
cd8c63889d
clarify documentation
2009-12-09 19:55:27 +00:00
Sears Russell
83a0ec68fe
fix rose bit rot, have unintialized page clear the impl pointer (for tidyness), remove some resolved @todos.
2009-12-04 19:33:09 +00:00
Sears Russell
7d1b702239
revert part of opensolaris patch--pack recordids again
2009-11-30 01:01:30 +00:00
Sears Russell
2ff421bff0
initial opensolaris port
2009-11-29 03:22:36 +00:00
Sears Russell
2535f69fbc
switch the concurrent hash to use lru->getStaleAndRemove; fix crashes due to partial leak of thread local pages on thread shutdown
2009-11-24 01:43:36 +00:00
Sears Russell
38b999dfed
add missing stlredblack.h header
2009-11-24 01:41:55 +00:00
Sears Russell
675bdce24a
wrote some wrappers that emulate redblack; default to stl tree for now
2009-11-24 00:24:18 +00:00
Sears Russell
1b0c10b37b
clean up handling of uninitialized pages (removes some special case code)
2009-11-09 23:02:27 +00:00
Sears Russell
795e044ab9
documentation fix
2009-11-09 22:58:37 +00:00
Sears Russell
b4a0f2436b
concurrent lru implementation for new buffer manager
2009-11-09 20:53:05 +00:00
Sears Russell
8bd0616f97
new concurrent buffer manager
2009-11-09 19:50:00 +00:00
Sears Russell
f5e3ab2d91
concurrent hashtable seems to be working; still need to work out atomicity primitives (eg; options to leave the bucket locked on success)
2009-10-19 21:55:25 +00:00
Sears Russell
1cf16f62f2
add optional support for gcc atomics to transaction table
2009-10-15 22:46:25 +00:00
Sears Russell
e6a718a7f9
improved concurrency for transaction table
2009-10-15 20:11:43 +00:00
Sears Russell
331f680c09
partial rewrite of transaction table to reduce thread contention
2009-10-15 02:33:08 +00:00
Sears Russell
4565aff9de
more refactoring; no more globals in transactionTable.c
2009-10-14 21:22:50 +00:00
Sears Russell
dbe3ecf0d0
refactoring transaction table; moved implementation into its own file. still need to remove globals
2009-10-14 18:57:50 +00:00
Sears Russell
20ef74b104
use pthread_rwlock instead of rwlock userspace emulation of it
2009-10-13 00:29:35 +00:00
Sears Russell
20a3fe9df5
add support for soft commits, and making prior soft commits durable
2009-10-12 19:16:41 +00:00
Sears Russell
942ba7fffc
bugfixes + support for hard limits on log size
2009-10-08 22:48:58 +00:00
Sears Russell
6a02f2f159
buffer manager implementation is now determined by an overridable factory method
2009-10-05 22:39:09 +00:00
Sears Russell
24650fa190
buffer manager refactoring; started work on new dirty page table writeback interface (to avoid circular dependencies)
2009-10-05 21:19:01 +00:00
Sears Russell
4689582cb5
improve log performance by disabling O_SYNC by default (and use fdatasync instead)
2009-09-12 16:20:13 +00:00
Sears Russell
c4e80b41b5
more include cleaning on alloc, begin cleaning of bufferHash
2009-08-22 21:29:27 +00:00
Sears Russell
1ce1496226
dependency cleanups; removed compenastions from alloc.c
2009-08-22 00:01:02 +00:00
Sears Russell
94c542f8b1
Increase dirtyPageTable concurrency
...
Add special-casing of lhtable; special case for buffer manager.
2009-08-14 06:31:13 +00:00
Sears Russell
dd0bcae074
add thread synchronization logic to ensure that only one thread is performing writeback at a time. This fixes weird I/O interleavings on some machines
2009-08-13 23:18:49 +00:00
Sears Russell
0154f7d29d
IMPORTANT API CHANGE: stasis_dirty_page_table_flush_range() no longer force writes to disk. You now need to call forcePageRange() after flush page range returns.
...
This commit should significantly improve sequential write thoughput.
I found a segfault bug this morning, but cannot reproduce. Could be a heisenbug, or could have been a hardware fault (gojira repeatedly kernel paniced, then both problems went away...)
2009-08-13 05:53:48 +00:00
Sears Russell
0cc65aefaf
Remove LRU-ordered writeback. Instead, writeback pages sequentially once some fraction of the page file is dirty. Also. further simplifies writeback code.
2009-08-11 17:23:43 +00:00
Sears Russell
907a24b68e
dirty page table flush was pinning pages, then asserting they were not pinned; updated page writeback api to allow correct usage
2009-08-08 01:43:03 +00:00
Sears Russell
c92ee87c32
improved behavior for sequential I/O performance, fixed races during writeback
2009-08-07 18:27:52 +00:00
Sears Russell
a6aa8ed942
added method to access dirtyPageTable
2009-08-05 20:02:33 +00:00
Sears Russell
5b10bcc63b
Reimplemented allocationPolicy.c; updated API + added more cases to the unit tests.
2009-08-05 06:35:11 +00:00
Sears Russell
c961b34903
forgot to commit b-tree source files
2009-07-29 21:04:08 +00:00
Sears Russell
ebd6b14dbf
single-page btrees are working
2009-07-29 21:03:16 +00:00
Sears Russell
0a308f035c
more header cleaning; updated INSTALL
2009-07-26 19:47:05 +00:00
Sears Russell
081c61b414
cleaned up includes, autoconf is now optional, since cmake now creates a config.h
2009-07-26 18:51:45 +00:00
Sears Russell
260310e151
Add page ops for tree implementations.
2009-07-26 06:18:54 +00:00
Sears Russell
7f2edd880d
New (more thorough, less frequent) tests for slotted pages. Renamed slotted methods, added microbenchmark for slotted pages.
2009-07-20 17:42:43 +00:00
Sears Russell
9ef3adf40c
- Fix compile warning + broken fcn pointer from last commit.
...
- Address crashes in Tupdate() and recovery caused by mixing segments and page oriented recovery (add loadPageForOperation() call)
2009-07-16 03:05:32 +00:00
Sears Russell
b44f8b17b3
Add getCachedPage() call. This allows dirtyPageTable to write back pages with out accidentally reading them back in from disk with loadPage(). This should improve performance and allow loadPageOfType() to be used safely.
2009-07-14 07:46:47 +00:00
Sears Russell
4493dbb88b
Added support for segment based recovery. For now, Stasis doesn't support hybrid recovery, so mixing page deallocation and segments will make recovery corrupt data / segfault.
2009-07-13 17:18:01 +00:00
Sears Russell
d6c91241d6
fixed bug that caused loadPageOfType to ignore type parameter. Recovery is still broken, since it calls loadPage, even when it can't
2009-07-10 01:19:44 +00:00
Sears Russell
eb0661ceeb
added segment.c
2009-07-09 21:09:36 +00:00
Sears Russell
ade1dc896e
Add loadPageOfType() call. (Untested for now)
2009-07-07 21:33:49 +00:00
Sears Russell
cf49c7dbf1
forgot to svn rm indirect's files
2009-07-07 20:05:15 +00:00
Sears Russell
2593de7a2c
Move page type to Page_s struct. This breaks indirect.c (which is poorly thought out, and will be removed shortly), and was the next step on the road to segment-based recovery
2009-07-07 19:20:22 +00:00
Sears Russell
b66328192c
in-memory group by implementation
2009-07-01 01:33:24 +00:00
Sears Russell
5efebbbb10
fix doxygen warning
2009-06-28 19:01:13 +00:00
Sears Russell
09af29abad
clean up iterator interface; remove Titerator_releaseLock()
2009-06-28 18:54:57 +00:00
Sears Russell
120665ae53
cleaned up arrayList. stasis_record_write no longer takes an LSN (it was ignoring it anyway). Added some missing calls to stasis_page_lsn_write (each time a page is written to, stasis_page_lsn_write should be called; this happens automatically for operations called via tupdate
2009-06-02 18:25:35 +00:00
Sears Russell
25c6be64b4
add assertunlocked() call
2009-05-30 02:22:04 +00:00
Sears Russell
7db06197ca
refactored truncation + dirty page table, fixed doxygen warnings.
2009-05-20 21:23:51 +00:00
Sears Russell
73c4b4fb71
Fix for Issue 1. Reworked log format so that nested top actions do not need to look up old entries during forward operation. This saves log bandwidth and avoids the bug that caused the performance regression
2009-05-14 21:28:09 +00:00
Sears Russell
3ff75ff568
made pass over log documentation
2009-05-13 22:06:58 +00:00
Sears Russell
87218ad929
yet another refactoring; broke groupForce into its own module, removed more static variables and namespace violations
2009-05-13 18:04:53 +00:00
Sears Russell
bc554efc85
renamed some functions to match naming convention
2009-05-08 06:53:30 +00:00
Sears Russell
651a1a22e5
more refactoring; removed static variables from pageHandle.c
2009-05-08 04:56:34 +00:00
Sears Russell
221b2e4acb
remove static variables, fix filehandle leak
2009-05-07 09:56:36 +00:00
Sears Russell
7b4cf40221
refactored/created stasis_buffer_manager_open and stasis_handle_open
2009-05-07 08:52:06 +00:00
Sears Russell
b5b414490a
fix config.h error for installed copies of stasis
2009-04-29 21:31:39 +00:00
Sears Russell
55c43358b6
fix recovery bug due to interactions between truncation and the old CLR format. check_faultInjection should now pass.
2009-04-22 22:03:38 +00:00
Sears Russell
731da7456a
more eclipse refactoring; cleaned up some names, whitespace and documentation
2009-04-13 22:50:12 +00:00
Sears Russell
7db35ecd39
removed static variables from bufferPool.c; renamed methods, moved Page typedef to common.h
2009-04-13 21:55:08 +00:00
Sears Russell
4b07b538a6
Renamed some methods; fixed bug that caused recovery to create potentially unbounded numbers of concurrent, active transactions.
...
(Note: this commit contains quite a few instances of auto-cleaned whitespace....)
2009-04-11 17:17:42 +00:00
Sears Russell
05077b62fd
nested top action cleanup and bugfix; subsequent recovery calls could undo logical undos in some circumstances
2009-04-03 22:01:37 +00:00