From: Daevid Vincent
Date: July 6 2006 7:17pm
Subject: RE: PBXT version 0.9.5 has been released
List-Archive: http://lists.mysql.com/mysql/199700
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Your site has a bunch of JS errors (using IE) so I can't roll over ANY =
of
the menus (left or upper right). I also cannot write to 'contact' =
because of
this same error. Hence I send it here... To the list... *sigh*

Can I also suggest not using a dark red hyperlink with black text. I =
didn't
even realize those were links until I randomly rolled over one and it
highlighted bright red.

D=C6VID =20

> -----Original Message-----
> From: Paul McCullagh [mailto:paul.mccullagh@stripped]=20
> Sent: Wednesday, July 05, 2006 6:51 AM
> To: mysql@stripped
> Subject: PBXT version 0.9.5 has been released
>=20
> Hi All,
>=20
> PBXT version 0.9.5 has been released and can be downloaded from=20
> http://www.primebase.com/xt.
>=20
> This version includes a number of major structural changes to the=20
> implementation in order to improve performance. In addition, all bugs=20
> reported so far have been fixed. For further information on this=20
> release, please read my blog, http://pbxt.blogspot.com, and the=20
> associated comments. Details of all changes made are given in the=20
> release notes below.
>=20
> PrimeBase XT (PBXT) is a new transactional database engine for MySQL.=20
> It has been designed for modern, web-based, high concurrency=20
> environments. The Beta release of PBXT is planned for September 2006.=20
> The current version is based on MySQL 4.1.16 and runs on Mac OS X and=20
> Linux. The next version will include support for the new 5.1 storage=20
> engine API and 64-bit processors.
>=20
> Any questions, comments and bug reports can be sent directly to me.=20
> Thanks for your support!
>=20
> Best regards,
>=20
> Paul McCullagh
> SNAP Innovation GmbH
>=20
>=20
> PBXT Release Notes
> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
>=20
> ------- 0.9.5 - 2006-07-03
>=20
> RN15: This version concludes the re-structuring of the PBXT=20
> implementation. I have made a number of major changes, including:
>=20
> - All files except the transaction logs are now associated with a=20
> particular table. All table related files begin with the name of the=20
> table. The extension indicates the function.
>=20
> - I have merged the handle and the fixed length row data for=20
> performance reasons.
>=20
> - Only the variable size component of a row is stored in the data log=20
> files. As a result the data logs can now be considered as a type of=20
> "overflow" area.
>=20
> - Memory mapped files are no longer used because it is not=20
> possible to=20
> flush changes to the disk.
>=20
> RN14: File names have the following forms:
>=20
> [table-name]-[table-id].xtr - These files contains the table row=20
> pointers. Each row pointer occupies 8 bytes and refers to a list of=20
> records. The file name also contains the table ID. This is a unique=20
> number which is used internally by XT to identify the table.
>=20
> [table-name].xtd - This file contains the fixed length data=20
> of a table.=20
> Each data item includes a handle and a record. The handle=20
> references a=20
> record in the data log file if the table contains variable length=20
> records.
>=20
> [table-name].xti - This file contains the index data of the table.
>=20
> [table-name]-[log-id].xtl - This is a data log file. It contains the=20
> variable length data of the table. A table may have any=20
> number of data=20
> log files, each with a unique ID.
>=20
> xtlog-[log-id].xt - These files are the transaction logs. Log entries=20
> that specify updates reference a data file record. Each active thread=20
> has its own transaction log in order to avoid contension.
>=20
> RN13: Fixed the bug "Hang on DROP DATABASE". [RESOLVED: 0016]
>=20
> RN12: PBXT currently only supports the "Serializable" transaction=20
> isolation level. This is the highest isolation level possible and=20
> includes the "repeatable-read" functionality [RESOLVED:=20
> 0015]. This is=20
> implemented by giving every transaction a snapshot of the database at=20
> the point when the transaction is started.
>=20
> If the transaction tries to update a record that was updated by some=20
> other transaction after the snapshot was taken, a locked error is=20
> returned. A deadlock can occur if 2 transactions update the=20
> same record=20
> in a different order. PBXT can detect all deadlocks.
>=20
> RN11: I have implemented write buffering on the table data files.=20
> [RESOLVED: 0013]
>=20
> RN10: The unique constraint (UNIQUE INDEX/PRIMARY KEY) is now checked=20
> correctly. [RESOLVED: 0008]
>=20
> RN9: I have implemented a conventional B-tree algorithm for=20
> the indices=20
> (instead of the Lehman and Yoa B*-link tree). Although this reduces=20
> concurrency it improves the performance of queries significantly=20
> because of the simplicity of the algorithm. Deletion is also=20
> implemented in a very simple manner. [RESOLVED: 0007]
>=20
> RN8: PBXT now has only 2 caches [RESOLVED: 0006]:
>=20
> The Index Cache (pbxt_index_cache_size): This is the amount of memory=20
> the PBXT storage engine uses to cache index data and row=20
> pointers. This=20
> is all the data in the files with the extensions '.xti' and '.xtr'.=20
> This cache is managed in blocks of 2K.
>=20
> The Record Cache (pbxt_record_cache_size): This is the amount=20
> of memory=20
> the PBXT storage engine uses to cache table row data (handles and=20
> records). This is all the data in the files with the extension '.xtd'.
>=20
> The size of the caches are determined by the values of the system=20
> variables pbxt_index_cache_size and pbxt_row_cache_size. By default=20
> these values are set to 32MB.
>=20
> RN7: Auto-increment is now implemented in memory. This is=20
> done by doing=20
> a MAX() select when a table is first opened to get the high value.=20
> After that, then high value is incremented in memory on INSERT. On=20
> UPDATE (or INSERT) the value in memory is adjusted if necessary. This=20
> method also makes it possible for rows to be inserted=20
> simultaneously on=20
> the same table. [RESOLVED: 0005, 0014]
>=20
> RN6: ./run-all-tests --create-options=3DTYPE=3DPBXT succeeds. =
[RESOLVED:=20
> 0004, 0019]
>=20
> RN5: Using sql-bench and my own Java based test I have confirmed that=20
> PBXT behaves correctly during multi-threaded access. [PARTIALY=20
> RESOLVED: 0002]
>=20
> RN4: Load/Stability test. Using sql-bench I have tested PBXT=20
> under load=20
> over a long period of time. [PARTIALY RESOLVED: 0001]
>=20
> ------- 0.9.2 - 2006-04-01
>=20
> RN3: Fixed a bug that cause the error "-6: Handle is out of range:=20
> [0:0]".
>=20
> RN2: Implemented SET, ENUM and YEAR data types.
>=20
> RN1: Fixed a bug in the error reporting when a table is=20
> created with a=20
> datatype that is not supported. [RESOLVED: 0011]
>=20
>=20
> --=20
> MySQL General Mailing List
> For list archives: http://lists.mysql.com/mysql
> To unsubscribe:   =20
> http://lists.mysql.com/mysql?unsub=3Ddaevid@stripped
>=20
>=20