2.1.6 is out! It was not planned, but there was a regression in 2.1.5 which caused the file to grow exponentially whenever a low cache size was configured. 2.1.6 fixes this regression.
Also in 2.1.6 is an improvement in the logging/journalling code base: both were merged, and the files are now buffered. This increases performance and scalability, but obviously you will only benefit if you use Transactions or if you enabled Recovery.
Everything is available for download here: http://hamsterdb.com/download.
chris Coding, Databases, hamsterdb, Libraries
This is the first release of the new Erlang driver. It supports most of the functionality, including Transactions and Cursors.
Sources are available on github, building them is straight forward.
http://github.com/cruppstahl/hamsterdb-erlang.
chris Coding, Databases, hamsterdb, Libraries
The new version 2.1.5 has a couple of bug fixes which crept in during 2.1.4. The bugs came to surface when I extended the ham_bench test tool with several more options. In addition, a few minor bugs were found with QuviQ‘s QuickCheck – an Erlang based software for property based testing. (BTW: Erlang support is on its way, will be released soon.)
In addition, several fixes improved performance and scalability. The freelist was replaced as well – it was a major obstacle for improved scaling.
One of the fixed bugs existed in older versions (pre 2.1.4): the cache was sometimes not purged. hamsterdb therefore required more memory than it should. This bug was fixed, but the fix *can* lead to performance regressions because the cache is now purged more frequently. This regression happens if your working set does not fit in the cache, obviously. A simple fix would be to increase the cache size.
See the README and the ChangeLog for more information. Files can be downloaded here: http://hamsterdb.com/download.
chris Databases, hamsterdb, Libraries
2.1.4 is out!
It features more Btree improvements: custom memory layouts for variable length keys and for duplicate keys. This layout stores a small up-front index in each node, with offsets to the key data and the record id (or inline record data). No overflow blobs are required for the key data, and since the keys no longer have fixed size slots there’s no memory wasted for keys which are smaller than this slot.
Records can now be stored directly in the leaf – this will be done automatically if the record is small enough and a certain number of key/record pairs fits into each page.
I’m now working on scalability improvements and thinking about rewriting parts of the file format. I hope that after 2.1.5 the file format will be stable again.
This release’s file format is not compatible to 2.1.3.
Everything’s available for download here: http://hamsterdb.com/download.
chris Coding, Databases, hamsterdb
Same procedure as every year, Mrs. Sophie?
Just in time for christmas, the server failed hard. Thanks for all the mails that everyone sent me. Since i’ve reached an age where i do no longer learn from my mistakes i had no backup. That’s why some posts are now missing.
Now that things finally work again i’ll focus on the next release; i hope it’ll be due in a week or so.
chris Uncategorized
This release fixes several issues including a segmentation fault, a few misplaced asserts and a data corruption. It is highly recommended to upgrade.
The number of serious issues that was reported to me over the last couple of weeks was quite astonishing, but on the other hand i had so much encouraging help from those who reported the bugs. Many helped in reproducing and they were really patient. Thanks everyone!
The most serious bug created a data corruption because the first freelist page in the database sometimes was not flushed to disk. This issue was not noticed over the last few years and my tests did not catch it. Now they do.
Everything’s available for download – including 130 megabytes of precompiled Win32/Win64 libraries (for MSVC 2008 and MSVC 2010).
http://www.hamsterdb.com/download
chris Coding, Databases, hamsterdb, Libraries
Next release – a tough one because i was swamped in work and really had troubles getting hamsterdb to run in Visual Studio 2010. What was supposed to be finished in 30 minutes then took about two weeks till it worked as expected.
Other changes include bug fixes at compilation time and link time, and a nasty bug which produced corrupt files: an important fsync was missing when flushing the log and the database file to disk. I hope that’s the last of these issues – they’re notoriously hard to test (and fix).
Files are available for download here: http://hamsterdb.com/download.
chris Coding, Databases, hamsterdb, Libraries
Just a quick note that i released 2.0.2; hamsterdb is now threadsafe.
It’s not yet concurrent and there’s only one big mutex for every Environment that you create. The next release will introduce a more fine-grained locking.
Also, a bug in the freelist was fixed therefore i recommend to update, even if you are not interested in thread-safety.
Boost is now required for building.
Files are available for download here: http://hamsterdb.com/download.
chris Coding, Databases, hamsterdb, Libraries
This release adds a few minor features like setting a custom path for log files and re-enabling approximate matching for use with Transactions. A few bugs were fixed as well.
New Features
Approximate matching now works if Transactions are enabled (only for ham_find)
Added new parameter HAM_PARAM_LOG_DIRECTORY to specify the directory where the log file and the journal files are stored
Bugfixes
Win32 release builds no longer crash when using Transactions
Win32: fixed build issue if HAM_DISABLE_COMPRESSION is defined (thanks, Andy Thalmann (http://www.softdev.ch))
hamsterdb 2.0.0 failed to load database files created with hamsterdb 1.x because of a bad version check
Fixed bug where HAM_ONLY_DUPLICATES was ignored when used in a Transaction (thanks, Mark Ruijter)
Other Changes
The internal ham_device_t structure was rewritten as a C++ class. If you use this structure to overwrite the device handling then you have to rewrite your code.
Now using the native Win32 API for os_pwrite and os_pread (thanks, Andy Thalmann (http://www.softdev.ch))
Added a public macro HAM_API_REVISION to allow users to programmatically distinguish between new and old interface of ham_txn_begin
Files are available for download here: http://hamsterdb.com/download.
chris Coding, Databases, hamsterdb, Libraries
After 1.5 years of work and 3 beta-releases I’m really happy to announce that 2.0.0 is finally stable!
Compared to rc3, a few minor performance improvements were added. Also, the interface of ham_txn_begin was changed. This is more or less the first API change since 1.0. I wrote a short migration guide.
Additionally, i added the Java and .NET APIs to the main repository. The Python API will follow with one of the next releases. This will basically force me to always keep them up-to-date. They were neglected for quite a long time.
And finally the webpage was redesigned. Have a look and tell me if you like it!
chris Coding, Databases, hamsterdb