A backend for ZODB that stores pickles in a relational database.
RelStorage is a storage implementation for ZODB that stores pickles in a relational database. PostgreSQL 8.1 and above (via psycopg2), MySQL 5.0.32+ / 5.1.34+ (via MySQLdb 1.2.2 and above), and Oracle 10g and 11g (via cx_Oracle) are currently supported. RelStorage replaced the PGStorage project.
- Included Utilities
- Migrating to a new version of RelStorage
- RelStorage Options
- Adapter Options
- Use with zodburi
- Project URLs
- Change History
- 1.6.0b1 (2014-09-04)
- 1.5.1 (2011-11-12)
- 1.5.0 (2011-06-30)
- 1.5.0b2 (2011-03-02)
- 1.5.0b1 (2011-02-05)
- 1.5.0a1 (2010-10-21)
- 1.4.2 (2011-02-04)
- 1.4.1 (2010-10-21)
- 1.4.0 (2010-09-30)
- 1.4.0c4 (2010-09-17)
- 1.4.0c3 (2010-07-31)
- 1.4.0c2 (2010-07-28)
- 1.4.0c1 (2010-06-19)
- 1.4.0b3 (2010-02-02)
- 1.4.0b2 (2010-01-30)
- 1.4.0b1 (2009-11-17)
- 1.3.0b1 (2009-09-04)
- 1.2.0 (2009-09-04)
- 1.2.0b2 (2009-05-05)
- It is a drop-in replacement for FileStorage and ZEO.
- There is a simple way to convert FileStorage to RelStorage and back again. You can also convert a RelStorage instance to a different relational database.
- Designed for high volume sites: multiple ZODB instances can share the same database. This is similar to ZEO, but RelStorage does not require ZEO.
- According to some tests, RelStorage handles high concurrency better than the standard combination of ZEO and FileStorage.
- Whereas FileStorage takes longer to start as the database grows due to an in-memory index of all objects, RelStorage starts quickly regardless of database size.
- Supports undo, packing, and filesystem-based ZODB blobs.
- Both history-preserving and history-free storage are available.
- Capable of failover to replicated SQL databases.
- Free, open source (ZPL 2.1)
You can install RelStorage using easy_install:
RelStorage requires a version of ZODB that is aware of MVCC storages. ZODB 3.9 supports RelStorage without any patches. ZODB 3.7 and 3.8 can support RelStorage if you first apply a patch to ZODB. You can get versions of ZODB with the patch already applied here:
The patches are also included in the source distribution of RelStorage.
You need the Python database adapter that corresponds with your database. Install psycopg2, MySQLdb 1.2.2+, or cx_Oracle 4.3+.
You need to configure a database and user account for RelStorage. RelStorage will populate the database with its schema the first time it connects.
If you installed PostgreSQL from a binary package, you probably have a user account named postgres. Since PostgreSQL respects the name of the logged-in user by default, switch to the postgres account to create the RelStorage user and database. Even root does not have the PostgreSQL privileges that the postgres account has. For example:
$ sudo su - postgres $ createuser --pwprompt zodbuser $ createdb -O zodbuser zodb
New PostgreSQL accounts often require modifications to pg_hba.conf, which contains host-based access control rules. The location of pg_hba.conf varies, but /etc/postgresql/8.4/main/pg_hba.conf is common. PostgreSQL processes the rules in order, so add new rules before the default rules rather than after. Here is a sample rule that allows only local connections by zodbuser to the zodb database:
local zodb zodbuser md5
PostgreSQL re-reads pg_hba.conf when you ask it to reload its configuration file:
Use the mysql utility to create the database and user account. Note that the -p option is usually required. You must use the -p option if the account you are accessing requires a password, but you should not use the -p option if the account you are accessing does not require a password. If you do not provide the -p option, yet the account requires a password, the mysql utility will not prompt for a password and will fail to authenticate.
Most users can start the mysql utility with the following shell command, using any login account:
$ mysql -u root -p
Here are some sample SQL statements for creating the user and database:
CREATE USER 'zodbuser'@'localhost' IDENTIFIED BY 'mypassword'; CREATE DATABASE zodb; GRANT ALL ON zodb.* TO 'zodbuser'@'localhost'; FLUSH PRIVILEGES;
Initial setup will require SYS privileges. Using Oracle 10g XE, you can start a SYS session with the following shell commands:
$ su - oracle $ sqlplus / as sysdba
You need to create a database user and grant execute privileges on the DBMS_LOCK package to that user. Here are some sample SQL statements for creating the database user and granting the required permissions:
CREATE USER zodb IDENTIFIED BY mypassword; GRANT CONNECT, RESOURCE, CREATE TABLE, CREATE SEQUENCE TO zodb; GRANT EXECUTE ON DBMS_LOCK TO zodb;
To install RelStorage in Plone, see the instructions in the following article:
Plone uses the plone.recipe.zope2instance Buildout recipe to generate zope.conf, so the easiest way to configure RelStorage in a Plone site is to set the rel-storage parameter in buildout.cfg. The rel-storage parameter contains options separated by newlines, with these values:
- type: any database type supported (postgresql, mysql, or oracle)
- RelStorage options like cache-servers and poll-interval
- Adapter-specific options
rel-storage = type mysql db plone user plone passwd PASSWORD
To integrate RelStorage in Zope 2, specify a RelStorage backend in etc/zope.conf. Remove the main mount point and add one of the following blocks. For PostgreSQL:
%import relstorage <zodb_db main> mount-point / <relstorage> <postgresql> # The dsn is optional, as are each of the parameters in the dsn. dsn dbname='zodb' user='username' host='localhost' password='pass' </postgresql> </relstorage> </zodb_db>
%import relstorage <zodb_db main> mount-point / <relstorage> <mysql> # Most of the options provided by MySQLdb are available. # See component.xml. db zodb </mysql> </relstorage> </zodb_db>
For Oracle (10g XE in this example):
%import relstorage <zodb_db main> mount-point / <relstorage> <oracle> user username password pass dsn XE </oracle> </relstorage> </zodb_db>
To add ZODB blob support, provide a blob-dir option that specifies where to store the blobs. For example:
%import relstorage <zodb_db main> mount-point / <relstorage> blob-dir ./blobs <postgresql> dsn dbname='zodb' user='username' host='localhost' password='pass' </postgresql> </relstorage> </zodb_db>
To use RelStorage with repoze.zodbconn, a package that makes ZODB available to WSGI applications, create a configuration file with contents similar to the following:
%import relstorage <zodb main> <relstorage> <mysql> db zodb </mysql> </relstorage> cache-size 100000 </zodb>
repoze.zodbconn expects a ZODB URI. Use a URI of the form zconfig://path/to/configuration#main.
RelStorage comes with a script named zodbconvert that converts databases between formats. Use it to convert a FileStorage instance to RelStorage and back, or to convert between different kinds of RelStorage instances, or to convert other kinds of storages that support the storage iterator protocol.
When converting between two history-preserving databases (note that FileStorage uses a history-preserving format), zodbconvert preserves all objects and transactions, meaning you can still use the ZODB undo feature after the conversion, and you can convert back using the same process in reverse. When converting from a history-free database to either a history-free database or a history-preserving database, zodbconvert retains all data, but the converted transactions will not be undoable. When converting from a history-preserving storage to a history-free storage, zodbconvert drops all historical information during the conversion.
Create a ZConfig style configuration file that specifies two storages, one named “source”, the other “destination”. The configuration file format is very much like zope.conf. Then run zodbconvert, providing the name of the configuration file as a parameter.
The utility does not modify the source storage. Before copying the data, the utility verifies the destination storage is completely empty. If the destination storage is not empty, the utility aborts without making any changes to the destination. (Adding transactions to an existing database is complex and out of scope for zodbconvert.)
Here is a sample zodbconvert configuration file:
<filestorage source> path /zope/var/Data.fs </filestorage> <relstorage destination> <mysql> db zodb </mysql> </relstorage>
This configuration file specifies that the utility should copy all of the transactions from Data.fs to a MySQL database called “zodb”. If you want to reverse the conversion, exchange the names “source” and “destination”. All storage types and storage options available in zope.conf are also available in this configuration file.
- Clears all data from the destination storage before copying. Use this only if you are certain the destination has no useful data. Currently only works when the destination is a RelStorage instance.
- Opens both storages and analyzes what would be copied, but does not actually copy.
RelStorage also comes with a script named zodbpack that packs any ZODB storage that allows concurrent connections (including RelStorage and ZEO, but not including FileStorage). Use zodbpack in cron scripts. Pass the script the name of a configuration file that lists the storages to pack, in ZConfig format. An example configuration file:
<relstorage> pack-gc true <mysql> db zodb </mysql> </relstorage>
- --days or -d
- Specifies how many days of historical data to keep. Defaults to 0, meaning no history is kept. This is meaningful even for history-free storages, since unreferenced objects are not removed from the database until the specified number of days have passed.
- Instructs the storage to only run the pre-pack phase of the pack but not actually delete anything. This is equivalent to specifying pack-prepack-only true in the storage options.
- Instructs the storage to only run the deletion (packing) phase, skipping the pre-pack analysis phase. This is equivalento to specifying pack-skip-prepack true in the storage options.
Sometimes RelStorage needs a schema modification along with a software upgrade. Hopefully, this will not often be necessary.
Migration to RelStorage version 1.5 requires a schema upgrade. See migrate-to-1.5.txt.
Migration to RelStorage version 1.4.2 requires a schema upgrade if you are using a history-free database (meaning keep-history is false). See migrate-to-1.4.txt.
See the notes subdirectory if you are upgrading from an older version.
Specify these options in zope.conf, as parameters for the relstorage.storage.RelStorage constructor, or as attributes of a relstorage.options.Options instance. In the latter two cases, use underscores instead of dashes in the option names.
- The name of the storage. Defaults to a descriptive name that includes the adapter connection parameters, except the database password.
- If true, only reads may be executed against the storage.
- If supplied, the storage will provide ZODB blob support; this option specifies the name of the directory to hold blob data. The directory will be created if it does not exist. If no value (or an empty value) is provided, then no blob support will be provided.
If true (the default), the blob directory is assumed to be shared among all clients using NFS or similar; blob data will be stored only on the filesystem and not in the database. If false, blob data is stored in the relational database and the blob directory holds a cache of blobs. When this option is false, the blob directory should not be shared among clients.
This option must be true when using ZODB 3.8, because ZODB 3.8 is not compatible with the file layout required for a blob cache. Use ZODB 3.9 or later if you want to store blobs in the relational database.
Maximum size of the blob cache, in bytes. If empty (the default), the cache size isn’t checked and the blob directory will grow without bounds. This should be either empty or significantly larger than the largest blob you store. At least 1 gigabyte is recommended for typical databases. More is recommended if you store large files such as videos, CD/DVD images, or virtual machine images.
This option allows suffixes such as “mb” or “gb”. This option is ignored if shared-blob-dir is true.
Blob cache check size as percent of blob-cache-size: “10” means “10%”. The blob cache size will be checked when this many bytes have been loaded into the cache. Defaults to 10% of the blob cache size.
This option is ignored if shared-blob-dir is true.
When ZODB blobs are stored in MySQL, RelStorage breaks them into chunks to minimize the impact on RAM. This option specifies the chunk size for new blobs. On PostgreSQL and Oracle, this value is used as the memory buffer size for blob upload and download operations. The default is 1048576 (1 mebibyte).
This option allows suffixes such as “mb” or “gb”. This option is ignored if shared-blob-dir is true.
If this option is set to true (the default), the adapter will create and use a history-preserving database schema (like FileStorage). A history-preserving schema supports ZODB-level undo, but also grows more quickly and requires extensive packing on a regular basis.
If this option is set to false, the adapter will create and use a history-free database schema. Undo will not be supported, but the database will not grow as quickly. The database will still require regular garbage collection (which is accessible through the database pack mechanism.)
This option must not change once the database schema has been installed, because the schemas for history-preserving and history-free storage are different. If you want to convert between a history-preserving and a history-free database, use the zodbconvert utility to copy to a new database.
If this option is provided, it specifies a text file that contains a list of database replicas the adapter can choose from. For MySQL and PostgreSQL, put in the replica file a list of host:port or host values, one per line. For Oracle, put in a list of DSN values. Blank lines and lines starting with # are ignored.
The adapter prefers the first replica specified in the file. If the first is not available, the adapter automatically tries the rest of the replicas, in order. If the file changes, the adapter will drop existing SQL database connections and make new connections when ZODB starts a new transaction.
Like the replica-conf option, but the referenced text file provides a list of database replicas to use only for read-only load connections. This allows RelStorage to load objects from read-only database replicas, while using read-write replicas for all other database interactions.
If this option is not provided, load connections will fall back to the replica pool specified by replica-conf. If ro-replica-conf is provided but replica-conf is not, RelStorage will use replicas for load connections but not for other database interactions.
Note that if read-only replicas are asynchronous, the next interaction after a write operation might not be up to date. When that happens, RelStorage will log a “backward time travel” warning and clear the ZODB cache to prevent consistency errors. This will likely result in temporarily reduced performance as the ZODB cache is repopulated.
- If this option has a nonzero value, when the adapter selects a replica other than the primary replica, the adapter will try to revert to the primary replica after the specified timeout (in seconds). The default is 600, meaning 10 minutes.
Specifies what to do when a database connection is stale. This is especially applicable to asynchronously replicated databases: RelStorage could switch to a replica that is not yet up to date.
When revert-when-stale is false (the default) and the database connection is stale, RelStorage will raise a ReadConflictError if the application tries to read or write anything. The application should react to the ReadConflictError by retrying the transaction after a delay (possibly multiple times.) Once the database catches up, a subsequent transaction will see the update and the ReadConflictError will not occur again.
When revert-when-stale is true and the database connection is stale, RelStorage will log a warning, clear the affected ZODB connection cache (to prevent consistency errors), and let the application continue with database state from an earlier transaction. This behavior is intended to be useful for highly available, read-only ZODB clients. Enabling this option on ZODB clients that read and write the database is likely to cause confusion for users whose changes seem to be temporarily reverted.
Defer polling the database for the specified maximum time interval, in seconds. Set to 0 (the default) to always poll. Fractional seconds are allowed. Use this to lighten the database load on servers with high read volume and low write volume.
The poll-interval option works best in conjunction with the cache-servers option. If both are enabled, RelStorage will poll a single cache key for changes on every request. The database will not be polled unless the cache indicates there have been changes, or the timeout specified by poll-interval has expired. This configuration keeps clients fully up to date, while removing much of the polling burden from the database. A good cluster configuration is to use memcache servers and a high poll-interval (say, 60 seconds).
This option can be used without the cache-servers option, but a large poll-interval without cache-servers increases the probability of basing transactions on stale data, which does not affect database consistency, but does increase the probability of conflict errors, leading to low performance.
If pack-gc is false, pack operations do not perform garbage collection. Garbage collection is enabled by default.
If garbage collection is disabled, pack operations keep at least one revision of every object. With garbage collection disabled, the pack code does not need to follow object references, making packing conceivably much faster. However, some of that benefit may be lost due to an ever increasing number of unused objects.
Disabling garbage collection is also a hack that ensures inter-database references never break.
- If pack-prepack-only is true, pack operations perform a full analysis of what to pack, but no data is actually removed. After a pre-pack, the pack_object, pack_state, and pack_state_tid tables are filled with the list of object states and objects that would have been removed. If pack-gc is true, the object_ref table will also be fully populated. The object_ref table can be queried to discover references between stored objects.
If pack-skip-prepack is true, the pre-pack phase is skipped and it is assumed the pack_object, pack_state and pack_state_tid tables have been filled already. Thus packing will only affect records already targeted for packing by a previous pre-pack analysis run.
Use this option together with pack-prepack-only to split packing into distinct phases, where each phase can be run during different timeslots, or where a pre-pack analysis is run on a copy of the database to alleviate a production database load.
- Packing occurs in batches of transactions; this specifies the timeout in seconds for each batch. Note that some database configurations have unpredictable I/O performance and might stall much longer than the timeout. The default timeout is 1.0 seconds.
- Before each pack batch, the commit lock is requested. If the lock is already held by for a regular commit, packing is paused for a short while. This option specifies how long the pack process should be paused before attempting to get the commit lock again. The default delay is 5.0 seconds.
Specifies a list of memcached servers. Using memcached with RelStorage improves the speed of frequent object accesses while slightly reducing the speed of other operations.
Provide a list of host:port pairs, separated by whitespace. “127.0.0.1:11211” is a common setting. Some memcached modules, such as pylibmc, allow you to specify a path to a Unix socket instead of a host:port pair.
The default is to disable memcached integration.
- Specifies which Python memcache module to use. The default is “relstorage.pylibmc_wrapper”, which requires pylibmc. An alternative module is “memcache”, a pure Python module. If you use the memcache module, use at least version 1.47. This option has no effect unless cache-servers is set.
- The prefix for all keys in the cache. All clients using a database should use the same cache-prefix. Defaults to the database name. (For example, in PostgreSQL, the database name is determined by executing SELECT current_database().) Set this if you have multiple databases with the same name.
- RelStorage caches pickled objects in memory, similar to a ZEO cache. The “local” cache is shared between threads. This option configures the approximate maximum amount of memory the cache should consume, in megabytes. It defaults to 10. Set to 0 to disable the in-memory cache.
- This option configures the maximum size of an object’s pickle (in bytes) that can qualify for the “local” cache. The size is measured before compression. Larger objects can still qualify for memcache. The default is 16384 (1 << 14) bytes.
- This option configures compression within the “local” cache. This option names a Python module that provides two functions, compress() and decompress(). Supported values include zlib, bz2, and none (no compression). The default is zlib.
- This is an advanced option. RelStorage uses a system of checkpoints to improve the cache hit rate. This option configures how many objects should be stored before creating a new checkpoint. The default is 10000.
During commit, RelStorage acquires a database-wide lock. This option specifies how long to wait for the lock before failing the attempt to commit. The default is 30 seconds.
The MySQL and Oracle adapters support this option. The PostgreSQL adapter currently does not.
- During commit, RelStorage acquires a database-wide lock. This option specifies the lock ID. This option currently applies only to the Oracle adapter.
- Normally, RelStorage will create or update the database schema on start-up. Set this option to false if you need to connect to a RelStorage database without automatic creation or updates.
The PostgreSQL adapter accepts:
Specifies the data source name for connecting to PostgreSQL. A PostgreSQL DSN is a list of parameters separated with whitespace. A typical DSN looks like:
dbname='zodb' user='username' host='localhost' password='pass'
If dsn is omitted, the adapter will connect to a local database with no password. Both the user and database name will match the name of the owner of the current process.
The MySQL adapter accepts most parameters supported by the MySQL-python library, including:
- string, host to connect
- string, user to connect as
- string, password to use
- string, database to use
- integer, TCP/IP port to connect to
- string, location of unix_socket (UNIX-ish only)
- mapping, maps MySQL FIELD_TYPE.* to Python functions which convert a string to the appropriate Python type
- number of seconds to wait before the connection attempt fails.
- if set, gzip compression is enabled
- if set, connect to server via named pipe (Windows only)
- command which is run once the connection is created
- see the MySQL documentation for mysql_options()
- see the MySQL documentation for mysql_options()
- client flags from MySQLdb.constants.CLIENT
- int, non-zero enables LOAD LOCAL INFILE, zero disables
The Oracle adapter accepts:
- The Oracle account name
- The Oracle account password
- The Oracle data source name. The Oracle client library will normally expect to find the DSN in /etc/oratab.
This package also enable the use of the postgres://, mysql:// and oracle:// URI schemes for zodburi. For more information about zodburi, please refer to its documentation. This section contains information specific to the these schemes.
The postgres:// , mysql:// and oracle:// URI schemes can be passed as zodbconn.uri to create a RelStorage PostgresSQL, MySQL or Oracle database factory. The uri should contain the user, the password, the host, the port and the db name e.g.:
Because oracle connection information are most often given as dsn, the oracle uri should not contain the same information as the other, but only the dsn
oracle://?dsn="HERE GOES THE DSN"
The URI scheme also accepts query string arguments. The query string arguments honored by this scheme are as follows.
Arguments that are usual with zodburi are also available here (see http://docs.pylonsproject.org/projects/zodburi/en/latest/) :
- boolean (if true, wrap RelStorage in a DemoStorage)
- integer (default 10000)
- integer (default 7)
An example that combines a path with a query string:
You can check out from Subversion using the following command:
svn co svn://svn.zope.org/repos/main/relstorage/trunk RelStorage
You can also browse the code:
The best place to discuss development of RelStorage is on the zodb-dev mailing list.
Q: How can I help improve RelStorage?
A: The best way to help is to test and to provide database-specific expertise. Ask questions about RelStorage on the zodb-dev mailing list.
Q: Can I perform SQL queries on the data in the database?
A: No. Like FileStorage and DirectoryStorage, RelStorage stores the data as pickles, making it hard for anything but ZODB to interpret the data. An earlier project called Ape attempted to store data in a truly relational way, but it turned out that Ape worked too much against ZODB principles and therefore could not be made reliable enough for production use. RelStorage, on the other hand, is much closer to an ordinary ZODB storage, and is therefore more appropriate for production use.
Q: How does RelStorage performance compare with FileStorage?
A: According to benchmarks, RelStorage with PostgreSQL is often faster than FileStorage, especially under high concurrency.
Q: Why should I choose RelStorage?
A: Because RelStorage is a fairly small layer that builds on world-class databases. These databases have proven reliability and scalability, along with numerous support options.
Q: Can RelStorage replace ZRS (Zope Replication Services)?
A: Yes, RelStorage inherits the asynchronous master/slave replication capability of MySQL and other databases. RelStorage has also been proven to work with Oracle RAC.
Q: How do I set up an environment to run the RelStorage tests?
A: See README.txt in the relstorage/tests directory.
- The local cache is now more configurable and uses zlib compression by default.
- Added support for zodburi, which means you can open a storage using “postgres:”, “mysql:”, or “oracle:” URIs.
- Packing: Reduced RAM consumption while packing by using IIBTree.Set instead of built-in set objects.
- MySQL 5.5: The test suite was freezing in checkBackwardTimeTravel. Fixed.
- Added performance metrics using the perfmetrics package.
- zodbconvert: Add an –incremental option to the zodbconvert script, letting you convert additional transactions at a later date, or update a non-live copy of your database, copying over missing transactions.
- Replication: Added the ro-replica-conf option, which tells RelStorage to use a read-only database replica for load connections. This makes it easy for RelStorage clients to take advantage of read-only database replicas.
- Replication: When the database connection is stale (such as when RelStorage switches to an asynchronous replica that is not yet up to date), RelStorage will now raise ReadConflictError by default. Ideally, the application will react to the error by transparently retrying the transaction, while the database gets up to date. A subsequent transaction will no longer be stale.
- Replication: Added the revert-when-stale option. When this option is true and the database connection is stale, RelStorage reverts the ZODB connection to the stale state rather than raise ReadConflictError. This option is intended for highly available, read-only ZODB clients. This option would probably confuse users of read-write ZODB clients, whose changes would sometimes seem to be temporarily reverted.
- Caching: Use the database name as the cache-prefix by default. This will hopefully help people who accidentally use a single memcached for multiple databases.
- Fixed compatibility with persistent 4.0.5 and above.
- Packing: Lowered garbage collection object reference finding log level to debug; this stage takes mere seconds, even in large sites, but could produce 10s of thousands of lines of log output.
- RelStorage was opening a test database connection (and was leaving it idle in a transaction with recent ZODB versions that support IMVCCStorage.) RelStorage no longer opens that test connection.
- zodbconvert: Avoid holding a list of all transactions in memory.
- Just after installing the database schema, verify the schema was created correctly. This affects MySQL in particular.
PostgreSQL: Fixed another minor compatibility issue with PostgreSQL 9.0. Packing raised an error when the client used old an version of libpq.
Delete empty transactions in batches of 1000 rows instead of all in one go, to prevent holding the transaction lock for longer than absolutely necessary.
Oracle: Fix object reference downloading performance for large RelStorage databases during the garbage collection phase of a pack.
Oracle, PostgreSQL: Switch to storing ZODB blob in chunks up to 4GB (the maximum supported by cx_Oracle) or 2GB (PostgreSQL maximum blob size) to maximize blob reading and writing performance.
The PostgreSQL blob_chunk schema changed to support this, see notes/migrate-to-1.5.txt to update existing databases.
zodbconvert: When copying a database containing blobs, ensure the source blob file exists long enough to copy it.
Better packing based on experience with large databases. Thanks to Martijn Pieters!
- Added more feedback to the packing process. It’ll now report during batch commit how much of the total work has been completed, but at most every .1% of the total number of transactions or objects to process.
- Renamed the –dry-run option to –prepack and added a –use-prepack-state to zodbpack. With these 2 options the pre-pack and pack phases can be run separately, allowing re-use of the pre-pack analysis data or even delegating the pre-pack phase off to a separate server.
- Replaced the packing duty cycle with a nowait locking strategy. The pack operation will now request the commit lock but pauses if it is already taken. It releases the lock after every batch (defaulting to 1 second processing). This makes the packing process faster while at the same time yielding to regular ZODB commits when busy.
- Do not hold the commit lock during pack cleanup while deleting rows from the object reference tables; these tables are pack-specific and regular ZODB commits never touch these.
Added an option to control schema creation / updating on startup. Setting the create-schema option to false will let you use RelStorage without a schema update.
Fixed compatibility with PostgreSQL 9.0, which is capable of returning a new ‘hex’ type to the client. Some builds of psycopg2 return garbage or raise an error when they see the new type. The fix was to encode more SQL query responses using base 64.
With the new shared-blob-dir option set to false, it was possible for a thread to read a partially downloaded blob. Fixed. Thanks for the report from Maurits van Rees.
Support for “shared-blob-dir false” now requires ZODB 3.9 or better. The code in the ZODB 3.8 version of ZODB.blob is not compatible with BlobCacheLayout, leading to blob filename collisions.
- Added a state_size column to object_state, making it possible to query the size of objects without loading the state. The new column is intended for gathering statistics. A schema migration is required.
- Added more logging during zodbconvert to show that something is happening and give an indication of how far along the process is.
- Fixed a missing import in the blob cache cleanup code.
- Added a –dry-run option to zodbpack.
- Replaced the graph traversal portion of the pack code with a more efficient implementation using Python sets (instead of SQL). The new code is much faster for packing databases with deeply nested objects.
- Added an option to store ZODB blobs in the database. The option is called “shared-blob-dir” and it behaves very much like the ZEO option of the same name. Blobs stored in the database are broken into chunks to reduce the impact on RAM.
- Require setuptools or distribute. Plain distutils is not sufficient.
- Fixed compatibility with ZODB 3.10. As reported by Jürgen Herrmann, there was a problem with conflict errors. The RelStorage patch of the sync() method now works with ZODB 3.10.
- Fixed a bug in packing history-free databases. If changes were made to the database during the pack, the pack code could delete too many objects. Thanks to Chris Withers for writing test code that revealed the bug. A schema migration is required for history-free databases; see notes/migration-to-1.4.txt.
- Enabled logging to stderr in zodbpack.
- Oracle: always connect in threaded mode. Without threaded mode, clients of Oracle 11g sometimes segfault.
- Made compatible with ZODB 3.10.0b7.
- Enabled ketama and compression in pylibmc_wrapper. Both options are better for clusters. [Helge Tesdal]
- Oracle: Use a more optimal query for POSKeyError logging. [Helge Tesdal]
- Fixed a NameError that occurred when getting the history of an object where transaction extended info was set. [Helge Tesdal]
- Worked around an Oracle RAC bug: apparently, in a RAC environment, the read-only transaction mode does not isolate transactions in the manner specified by the documentation, so Oracle users now have to use serializable isolation like everyone else. It’s slower but more reliable.
- Use the client time instead of the database server time as a factor in the transaction ID. RelStorage was using the database server time to reduce the need for synchronized clocks, but in practice, that policy broke tests and did not really avoid the need to synchronize clocks. Also, the effect of unsynchronized clocks is predictable and manageable: you’ll get bunches of transactions with sequential timestamps.
- If the database returns an object from the future (which should never happen), generate a ReadConflictError, hopefully giving the application a chance to recover. The most likely causes of this are a broken database and threading bugs.
- Always update the RelStorage cache when opening a database connection for loading, even when no ZODB Connection is using the storage. Otherwise, code that used the storage interface directly could cause the cache to fall out of sync; the effects would be seen in the next ZODB.Connection.
- Added a ZODB monkey patch that passes the “force” parameter to the sync method. This should help the poll-interval option do its job better.
- Fixed a subtle bug in the cache code that could lead to an AssertionError indicating a cache inconsistency. The inconsistency was caused by after_poll(), which was ignoring the randomness of the order of the list of recent changes, leading it to sometimes put the wrong transfer ID in the “delta_after” dicts. Also expanded the AssertionError with debugging info, since cache inconsistency can still be caused by database misconfiguration and mismatched client versions.
- Oracle: updated the migration notes. The relstorage_util package is not needed after all.
History-preserving storages now replace objects on restore instead of just inserting them. This should solve problems people were having with the zodbconvert utility.
Oracle: call the DBMS_LOCK.REQUEST function directly instead of using a small package named relstorage_util. The relstorage_util package was designed as a secure way to access the DBMS_LOCK package, but the package turned out to be confusing to DBAs and provided no real security advantage. People who have already deployed RelStorage 1.4.x on Oracle need to do the following:
GRANT EXECUTE ON DBMS_LOCK TO <zodb_user>;
You can also drop the relstorage_util package. Keep the relstorage_op package.
Made compatible with ZODB 3.10.
MySQL: specify the transaction isolation mode for every connection, since the default is apparently not necessarily “read committed” anymore.
- Auto-reconnect in new_oid().
- Include all test subpackages in setup.py.
- Raise an error if MySQL reverts to MyISAM rather than using the InnoDB storage engine.
- Added the keep-history option. Set it to false to keep no history. (Packing is still required for garbage collection and blob deletion.)
- Added the replica-conf and replica-timeout options. Set replica-conf to a filename containing the location of database replicas. Changes to the file take effect at transaction boundaries.
- Expanded the option documentation in README.txt.
- Revised the way RelStorage uses memcached. Minimized the number of trips to both the cache server and the database.
- Added an in-process pickle cache that serves a function similar to the ZEO cache.
- Added a wrapper module for pylibmc.
- Store operations now use multi-insert and multi-delete SQL statements to reduce the effect of network latency.
- Renamed relstorage.py to storage.py to overcome import issues. Also moved the Options class to options.py.
- Updated the patch for ZODB 3.7 and 3.8 to fix an issue with blobs and subtransactions.
- Divided the implementation of database adapters into many small objects, making the adapter code more modular. Added interfaces that describe the duties of each part.
- Oracle: Sped up restore operations by sending short blobs inline.
- Oracle: Use a timeout on commit locks. This requires installation of a small PL/SQL package that can access DBMS_LOCK. See README.txt.
- Oracle: Used PL/SQL bulk insert operations to improve write performance.
- PostgreSQL: use the documented ALTER SEQUENCE RESTART WITH statement instead of ALTER SEQUENCE START WITH.
- Moved MD5 sum computation to the adapters so they can choose not to use MD5.
- Changed loadSerial to load from the store connection only if the load connection can not provide the object requested.
- Stopped wrapping database disconnect exceptions. Now the code catches and handles them directly.
- Use the store connection rather than the load connection for OID allocation.
- Detect and handle backward time travel, which can happen after failover to an out-of-date asynchronous slave database. For simplicity, invalidate the whole ZODB cache when this happens.
- Replaced the speed test script with a separately distributed package, zodbshootout.
- Added the zodbpack script.
- Added support for a blob directory. No BlobStorage wrapper is needed. Cluster nodes will need to use a shared filesystem such as NFS or SMB/CIFS.
- Added the blob-dir parameter to the ZConfig schema and README.txt.
- In Oracle, trim transaction descriptions longer than 2000 bytes.
- When opening the database for the first time, don’t issue a warning about the inevitable POSKeyError on the root OID.
- If RelStorage tries to unpickle a corrupt object state during packing, it will now report the oid and tid in the log.
- RelStorage now implements IMVCCStorage, making it compatible with ZODB 3.9.0b1 and above.
- Removed two-phase commit support from the PostgreSQL adapter. The feature turned out to be unnecessary.
- Added MySQL 5.1.34 and above to the list of supportable databases.
- Fixed minor test failures under Windows. Windows is now a supportable platform.