summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
| | * ISAMTree: added comment explaining the compiler fenceDouglas Rumbaugh2024-02-081-0/+4
| | |
| | * Added compiler fence to block reorderingDouglas Rumbaugh2024-02-082-2/+4
| | | | | | | | | | | | I'm reasonably certain that this is a compiler bug...
| * | Added compiler fence to block reorderingDouglas Rumbaugh2024-02-082-2/+8
| |/ | | | | | | I'm reasonably certain that this is a compiler bug...
| * Fixed a slight synchronization bug in Epoch retirement "properly"Douglas Rumbaugh2024-02-081-10/+27
| |
| * Fixed benchmark memory leakDouglas Rumbaugh2024-02-081-0/+2
| |
| * Cleaned up shard files (except VPTree)Douglas Rumbaugh2024-02-0728-1674/+859
| | | | | | | | | | | | Cleaned up shard implementations, fixed a few bugs, and set up some tests. There's still some work to be done in creating tests for the weighted sampling operations for the alias and aug btree shards.
| * Fully realized shard concept interfaceDouglas Rumbaugh2024-02-0733-686/+77
| |
| * Fully implemented Query concept and adjusted queries to use itDouglas Rumbaugh2024-02-0713-56/+54
| |
| * BufferView: Adjusted BV to avoid repeated modulus operationsDouglas Rumbaugh2024-02-053-15/+190
| |
| * Set up tombstone deletes properlyDouglas Rumbaugh2024-02-058-56/+72
| |
| * updated btree benchmarkDouglas Rumbaugh2024-01-311-0/+4
| |
| * IRS bench (replication of existing one)Douglas Rumbaugh2024-01-311-0/+125
| |
| * BTree benchmarkDouglas Rumbaugh2024-01-313-6/+128
| |
| * Updated throughput bench to use SOSDDouglas Rumbaugh2024-01-312-21/+39
| |
| * temporary hack to get workingDouglas Rumbaugh2024-01-311-6/+12
| |
| * Adjusted epoch transition methodologyDouglas Rumbaugh2024-01-314-153/+121
| |
| * More locking!Douglas Rumbaugh2024-01-311-0/+2
| |
| * Updated insert query throughput to use IRS queriesDouglas Rumbaugh2024-01-314-123/+25
| |
| * insert query tput updatesDouglas Rumbaugh2024-01-311-2/+7
| |
| * Possibly fixed epoch retirement sync errorDouglas Rumbaugh2024-01-311-1/+5
| |
| * Insert throughput benchmarkDouglas Rumbaugh2024-01-313-46/+220
| |
| * Query/Insert intermix benchmarksDouglas Rumbaugh2024-01-314-41/+162
| |
| * Temporary thread affinity for reconstructionDouglas Rumbaugh2024-01-313-7/+40
| |
| * TrieSpline + testsDouglas Rumbaugh2024-01-304-301/+117
| |
| * Benchmarking updatesDouglas Rumbaugh2024-01-309-30/+495
| |
| * Multithreaded Insertion BenchmarkDouglas Rumbaugh2024-01-241-11/+37
| |
| * DynamicExtension: Fixed reconstruction trigger data raceDouglas Rumbaugh2024-01-241-4/+7
| | | | | | | | | | Tweak the reconstruction trigger code to ensure that multiple reconstructions won't be triggered at the same time.
| * Added watermark testing to CMakeDouglas Rumbaugh2024-01-221-0/+6
| |
| * WAtermark testing benchmarkDouglas Rumbaugh2024-01-221-0/+53
| |
| * Adjusted recon_benchmark and properly shutdown FIFOSchedulerDouglas Rumbaugh2024-01-222-12/+25
| |
| * Benchmarking: updated insert_query_tput to use better rngDouglas Rumbaugh2024-01-221-3/+14
| |
| * FIFOScheduler: added automake wakeupDouglas Rumbaugh2024-01-221-1/+12
| | | | | | | | | | | | | | | | | | | | | | | | Sometimes, when the max thread count is exceeded, it is possible for the scheduler to lock up. This is because the scheduler is only run when a new job is put into the queue, and so it is possible for a job to be blocked by thread limitations and be left sitting in the queue. If the main program is waiting on this job to finish before scheduling a new one, then the system deadlocks. I added a second background thread to the scheduler that wakes the scheduler up every 20us to resolve this and prevent these deadlocks.
| * Moved some benchmarks over to range countDouglas Rumbaugh2024-01-222-6/+6
| |
| * Added RangeCount queryDouglas Rumbaugh2024-01-224-0/+386
| |
| * Initial rough-out of internal statistics trackerDouglas Rumbaugh2024-01-198-36/+160
| | | | | | | | | | | | Need to figure out the best way to do the detailed tracking in a concurrent manner. I was thinking just an event log, with parsing routines for extracting statistics. But that'll be pretty slow.
| * Multiple concurrency bug fixesDouglas Rumbaugh2024-01-1714-185/+246
| | | | | | | | | | | | | | | | A poorly organized commit with fixes for a variety of bugs that were causing missing records. The core problems all appear to be fixed, though there is an outstanding problem with tombstones not being completely canceled. A very small number are appearing in the wrong order during the static structure test.
| * Concurrency testing and bug fixesDouglas B. Rumbaugh2024-01-156-8/+467
| |
| * Benchmarking programsDouglas B. Rumbaugh2024-01-1518-12/+132
| |
| * Use 16-byte CAS to control buffer headDouglas B. Rumbaugh2024-01-152-70/+77
| |
| * Various single-threaded bug fixesDouglas B. Rumbaugh2024-01-154-15/+55
| |
| * Initial integration of new buffering scheme into frameworkDouglas Rumbaugh2024-01-129-289/+158
| | | | | | | | | | It isn't working right now (lotsa test failures), but we're to the debugging phase now.
| * Initial pass at unit test refactoringDouglas Rumbaugh2024-01-1212-468/+757
| | | | | | | | | | | | Restructured unit tests to be a bit more modular. I have some further plans to expand on this, particular for the query tests (including both shard and framework level test functions that can be injected at will).
| * BufferView.h: Hopefully the last necessary tweak to the move semantics stuffDouglas Rumbaugh2024-01-121-24/+2
| | | | | | | | | | | | | | | | You can't move assign an std::Bind, but you can move construct it. So I had to disable the move assignment operator. This means that when you change the BufferView ownership over to, say, a QueryBufferState object, you need to do it by passing std::move(buffview) into a constructor call only--you cannot assign it.
| * InternalLevel update and testsDouglas Rumbaugh2024-01-114-51/+76
| | | | | | | | | | Plus some assorted fixes for move semantics stuff in BufferView that accompanied these changes.
| * Ported ISAMTree over to new buffer setupDouglas Rumbaugh2024-01-114-125/+137
| | | | | | | | | | I may still play with the shard from shards constructor, and queries need some work yet too.
| * BufferView: enforce move semanticsDouglas Rumbaugh2024-01-111-2/+19
| | | | | | | | | | | | Because a BufferView's lifetime is so tightly linked to the lifetime of regions of the buffer, it can't be copied without potentially breaking things.
| * Fixed some potential buffer-related concurrency bugsDouglas Rumbaugh2024-01-113-58/+95
| |
| * MutableBuffer: multithreaded insert test + bugfixesDouglas Rumbaugh2024-01-102-16/+29
| |
| * MutableBuffer+View: Implementation with unit testsDouglas Rumbaugh2024-01-103-190/+229
| |
| * Initial update of buffer to new specificationsDouglas B. Rumbaugh2024-01-092-152/+62
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There are a few minor issues that this introduces, however. Global tracking of a lot of secondary information, such as weights for WIRS/WSS, or the exact number of tombstones will need to be approached differently than they have been historically with this new approach. I've also removed most of the tombstone capacity related code. We had decided not to bother enforcing this at the buffer level anyway, and it would greatly increase the complexity of the problem of predicting when the next compaction will be. On the whole this new approach seems like it'll simplify a lot. This commit actually removes significantly more code than it adds. One minor issue: the currently implementation will have problems in the circular array indexes once more than 2^64 records have been inserted. This doesn't seem like a realistic problem at the moment.