Difference between revisions of "Transaction Management"

From BaseX Documentation
Jump to navigation Jump to search
(45 intermediate revisions by 5 users not shown)
Line 1: Line 1:
 
This article is part of the [[Advanced User's Guide]].
 
This article is part of the [[Advanced User's Guide]].
The BaseX client-server architecture offers ACID safe transactions,
+
The BaseX client-server architecture offers ACID-safe transactions,
with multiple readers and writers. Here are some more
+
with multiple readers and writers. Here is some more
informations about the transaction management.
+
information about the transaction management.
  
=Transaction=
+
=Introduction=
  
 
In a nutshell, a transaction is equal to a command or query. So each command or query sent to the server becomes a transaction.
 
In a nutshell, a transaction is equal to a command or query. So each command or query sent to the server becomes a transaction.
  
Incoming requests are parsed and checked for errors on the server. If the command or query is not correct, the request will not be executed,
+
Incoming requests are parsed and checked for errors on the server. If the command or query is not correct, the request will not be executed, and the user will receive an error message. Otherwise the request becomes a transaction and gets into the transaction monitor.
and the user will receive an error message. Otherwise the request becomes a transaction and gets into the transaction monitor.
 
  
Note:
+
Please note that:
An unexpected abort of the server during a transaction, caused by a hardware
 
failure or power cut, will probably lead to an inconsistent database state if a transaction was active at the shutdown time. So we advise to use
 
the [[Commands#BACKUP|BACKUP]] command to backup your database regularly. If the worst case occurs, you can try the [[Commands#INSPECT|INSPECT]] command to check if your database has obvious inconsistencies, and [[Commands#RESTORE|RESTORE]] to restore a previous version of the database.
 
  
==Update Transactions==
+
* Locks ''cannot be synchronized'' across BaseX instances that run in different JVMs. If concurrent write operations are to be performed, we generally recommend working with the client/server or the HTTP architecture .
 +
* An ''unexpected abort'' of the server during a transaction, caused by a hardware failure or power cut, may lead to an inconsistent database state if a transaction was active at shutdown time. So it is advisable to use the [[Commands#CREATE BACKUP|BACKUP]] command to regularly backup your database. If the worst case occurs, you can try the [[Commands#INSPECT|INSPECT]] command to check if your database has obvious inconsistencies, and use [[Commands#RESTORE|RESTORE]] to restore the last backed up version of the database.
 +
 
 +
==XQuery Update==
  
 
Many update operations are triggered by [[Update|XQuery Update]] expressions. When executing an updating query, all update operations of the query are stored in a pending update list. They will be executed all at once, so the database is updated atomically. If any of the update sub-operations is erroneous, the overall transaction will be aborted.
 
Many update operations are triggered by [[Update|XQuery Update]] expressions. When executing an updating query, all update operations of the query are stored in a pending update list. They will be executed all at once, so the database is updated atomically. If any of the update sub-operations is erroneous, the overall transaction will be aborted.
Line 22: Line 21:
 
=Concurrency Control=
 
=Concurrency Control=
  
BaseX 7.6 introduces locking on database level. Writing transactions do not necessarily block all other transactions any more. The number of parallel transactions can be limited by setting the [[Options#PARALLEL|PARALLEL]] option.
+
BaseX provides support for multiple read and single write operations (using preclaiming and starvation-free two phase locking). This means that:
 +
 
 +
* Read transactions are executed in parallel.
 +
* If an updating transaction comes in, it will be queued and executed after all previous read transaction have been executed.
 +
* Subsequent operations (read or write) will be queued until the updating transaction has completed.
 +
 
 +
Each database has its own queue: An update on database A will not block operations on database B. This is under the premise that it can be statically determined, i.e., before the transaction is evaluated, which databases will be accessed by a transaction (see [[#Limitations|below]]). The number of maximum parallel transactions can be adjusted with the [[Options#PARALLEL|PARALLEL]] option.
 +
 
 +
With {{Version|8.6}}, locking has been improved:
 +
 
 +
* Jobs without database access will never be locked. Globally locking jobs can now be executed in parallel with non-locking jobs.
 +
* A {{Option|FAIRLOCK}} option has been added: By default, read transactions will now be favored, and transactions that access no databases can be evaluated even if the transactions limit has been reached.
 +
 
 +
==External Side Effects==
 +
 
 +
Access to external resources (files on hard disk, HTTP requests, ...) is not controlled by the transaction monitor of BaseX unless specified by the user.
 +
 
 +
===XQuery Locking Options===
 +
 
 +
Custom locks can be acquired by setting the BaseX-specific XQuery options {{Code|query:read-lock}} and {{Code|query:write-lock}}. Multiple option declarations may occur in the prolog of a query, but multiple values can also be separated with commas in a single declaration. These locks are in another namespace than the database names: the lock value {{Code|factbook}} will not lock a database named factbook.
  
==Transaction Monitor==
+
These option declarations will put read locks on ''foo'', ''bar'' and ''batz'' and a write lock on ''quix'':
  
The transaction monitor ensures that just one writing transaction or an arbitrary amount of reading transactions ''per database'' are active at the same time.
+
<pre class="brush:xquery">
 +
declare option query:read-lock "foo,bar";
 +
declare option query:read-lock "batz";
 +
declare option query:write-lock "quix";
 +
</pre>
  
Deadlocks are prevented by using preclaiming two phase locking. Execution is starvation-free as lock aquiration is queued per database. Due to the specifics of XQuery Update, all updates are written at the end of the query. Locking is strict with the exception that databases for which BaseX recognizes it will not write to are downgraded to read locks.
+
===Java Modules===
  
Locks are not synchronized between multiple BaseX instances. We generally recommend working with the client/server architecture if concurrent write operations are to be performed.
+
Locks can also be acquired on [[Java Bindings#Locking|Java functions]] which are imported and invoked from an XQuery expression. It is advisable to explicitly lock Java code whenever it performs sensitive read and write operations.
  
 
==Limitations==
 
==Limitations==
Line 36: Line 58:
 
===Commands===
 
===Commands===
  
Database locking works with all commands unless no glob syntax is used, such as in the following command call:
+
Database locking works with all commands unless the glob syntax is used, such as in the following command call:
  
 
* {{Code|DROP DB new*}}: drop all databases starting with "new"
 
* {{Code|DROP DB new*}}: drop all databases starting with "new"
Line 42: Line 64:
 
===XQuery===
 
===XQuery===
  
As XQuery is a very powerful language, deciding which databases will be accessed by a query is non-trivial. Optimization is work in progress.
+
Deciding which databases will be accessed by a complex XQuery expression is a non-trivial task. Database detection works for the following types of queries:
The current identification of which databases to lock is limited to queries that access the currently opened database, XQuery functions that explicitly specify a database, and expressions that address no database at all.
 
 
 
Some examples on database-locking enabled queries:
 
  
 
* {{Code|//item}}, read-locking of the database opened by a client
 
* {{Code|//item}}, read-locking of the database opened by a client
Line 53: Line 72:
 
* {{Code|delete nodes doc('test')//*[string-length(local-name(.)) > 5]}}, write-locking of "test"
 
* {{Code|delete nodes doc('test')//*[string-length(local-name(.)) > 5]}}, write-locking of "test"
  
All of these queries can be executed in parallel:
+
All databases will be locked by queries of the following kind:
  
Some examples on queries that are not supported by database-locking yet:
+
* {{Code|for $db in ('db1', 'db2') return doc($db)}}
 +
* {{Code|doc(doc('test')/reference/text())}}
 +
* <code>let $db := 'test' return insert nodes <test/> into doc($db)</code>
  
* <code>let $db := 'factbook' return doc($db)</code>, will read lock globally: referencing database names isn’t supported yet
+
You can consult the query info output (which you find in the [[GUI#Visualizations|Info View]] of the GUI or which you can turn on by setting <code>[[Options#QUERYINFO|QUERYINFO]]</code> to {{Code|true}}) to find out which databases have been locked by a query.
* {{Code|for $db in ('factbook') return doc($db)}}, will read lock globally: ditto
 
* {{Code|doc(doc('test')/reference/text())}}, will read lock globally: ditto
 
* <code>let $db := 'test' return insert nodes <test/> into doc($db)</code>, will write lock globally: ditto
 
 
 
A list of all locked databases is output if <code>[[Options#QUERYINFO|QUERYINFO]]</code> is set to {{Code|true}}. <!-- and in the GUI's [[GUI#Visualizations|Info View]] --> If you think too much is locked, please give us a note on our [http://basex.org/open-source/ mailing list] with some example code.
 
 
 
===GUI===
 
 
 
Database locking is currently disabled if the BaseX GUI is used.
 
 
 
==How to disable==
 
 
 
In order to enable traditional process locking, the option <code>[[Options#GLOBALLOCK|GLOBALLOCK]]</code> can be set to {{Code|true}}. This can e.g. be done by editing your {{Code|.basex}} file (see [[Options]] for more details).
 
  
 
=File-System Locks=
 
=File-System Locks=
Line 76: Line 84:
 
==Update Operations==
 
==Update Operations==
  
During the term of a database update, a locking file {{Code|upd.basex}} will reside in that database directory. If the update fails for some unexpected reason, or if the process is killed ungracefully, this file may not be deleted. In this case, the database cannot be opened anymore using the default commands, and the message "Database ... is being updated, or update was not completed" will be shown instead. If the locking file is manually removed, you may be able to reopen the database, but you should be aware that database may have got corrupt due to the interrupted update process, and you should revert to the most recent database backup.
+
During a database update, a locking file {{Code|upd.basex}} will reside in that database directory. If the update fails for some unexpected reason, or if the process is killed ungracefully, this file will not be deleted. In this case, the database cannot be opened anymore, and the message "Database ... is being updated, or update was not completed" will be shown instead.  
 +
 
 +
If the locking file is manually removed, you may be able to reopen the database, but you should be aware that database may have got corrupt due to the interrupted update process, and you should revert to the most recent database backup.
  
 
==Database Locks==
 
==Database Locks==
  
To avoid database corruptions caused by write operations running in different JVMs, a shared lock is requested on the database table file ({{Code|tbl.basex}}) whenever a database is opened. If an update operation is triggered, it will be rejected with the message "Database ... is opened by another process." if no exclusive lock can be acquired.
+
To avoid database corruptions that are caused by accidental write operations from different JVMs, a shared lock is requested on the database table file ({{Code|tbl.basex}}) whenever a database is opened. If an update operation is triggered, and if no exclusive lock can be acquired, it will be rejected with the message "Database ... is currently opened by another process.".
  
As the standalone versions of BaseX (command-line, GUI) cannot be synchronized with other BaseX instances, we generally recommend working with the client/server architecture if concurrent write operations are to be performed.
+
Please note that you cannot 100% rely on this mechanism, as it is not possible to synchronize operations across different JVMs. You will be safe when using the client/server or HTTP architecture.
  
 
=Changelog=
 
=Changelog=
 +
 +
;Version 8.6
 +
* Updated: New {{Option|FAIRLOCK}} option, improved detection of lock patterns.
 +
 +
;Version 7.8
 +
* Added: Locks can also be acquired on [[Java Bindings#Locking|Java functions]].
  
 
;Version 7.6
 
;Version 7.6
 
+
* Added: database locking introduced, replacing process locking.
* Added: database locking introduced, replacing process locking
 
  
 
;Version 7.2.1
 
;Version 7.2.1
 
+
* Updated: pin files replaced with shared/exclusive filesystem locking.
* Updated: pin files replaced with shared/exclusive filesystem locking
 
  
 
;Version 7.2
 
;Version 7.2
 
+
* Added: pin files to mark open databases.
* Added: pin files to mark open databases
 
  
 
;Version 7.1
 
;Version 7.1
 
+
* Added: update lock files.
* Added: update lock files
 
 
 
[[Category:Server]]
 
[[Category:Internals]]
 

Revision as of 18:35, 24 January 2017

This article is part of the Advanced User's Guide. The BaseX client-server architecture offers ACID-safe transactions, with multiple readers and writers. Here is some more information about the transaction management.

Introduction

In a nutshell, a transaction is equal to a command or query. So each command or query sent to the server becomes a transaction.

Incoming requests are parsed and checked for errors on the server. If the command or query is not correct, the request will not be executed, and the user will receive an error message. Otherwise the request becomes a transaction and gets into the transaction monitor.

Please note that:

  • Locks cannot be synchronized across BaseX instances that run in different JVMs. If concurrent write operations are to be performed, we generally recommend working with the client/server or the HTTP architecture .
  • An unexpected abort of the server during a transaction, caused by a hardware failure or power cut, may lead to an inconsistent database state if a transaction was active at shutdown time. So it is advisable to use the BACKUP command to regularly backup your database. If the worst case occurs, you can try the INSPECT command to check if your database has obvious inconsistencies, and use RESTORE to restore the last backed up version of the database.

XQuery Update

Many update operations are triggered by XQuery Update expressions. When executing an updating query, all update operations of the query are stored in a pending update list. They will be executed all at once, so the database is updated atomically. If any of the update sub-operations is erroneous, the overall transaction will be aborted.

Concurrency Control

BaseX provides support for multiple read and single write operations (using preclaiming and starvation-free two phase locking). This means that:

  • Read transactions are executed in parallel.
  • If an updating transaction comes in, it will be queued and executed after all previous read transaction have been executed.
  • Subsequent operations (read or write) will be queued until the updating transaction has completed.

Each database has its own queue: An update on database A will not block operations on database B. This is under the premise that it can be statically determined, i.e., before the transaction is evaluated, which databases will be accessed by a transaction (see below). The number of maximum parallel transactions can be adjusted with the PARALLEL option.

With Version 8.6, locking has been improved:

  • Jobs without database access will never be locked. Globally locking jobs can now be executed in parallel with non-locking jobs.
  • A FAIRLOCK option has been added: By default, read transactions will now be favored, and transactions that access no databases can be evaluated even if the transactions limit has been reached.

External Side Effects

Access to external resources (files on hard disk, HTTP requests, ...) is not controlled by the transaction monitor of BaseX unless specified by the user.

XQuery Locking Options

Custom locks can be acquired by setting the BaseX-specific XQuery options query:read-lock and query:write-lock. Multiple option declarations may occur in the prolog of a query, but multiple values can also be separated with commas in a single declaration. These locks are in another namespace than the database names: the lock value factbook will not lock a database named factbook.

These option declarations will put read locks on foo, bar and batz and a write lock on quix:

declare option query:read-lock "foo,bar";
declare option query:read-lock "batz";
declare option query:write-lock "quix";

Java Modules

Locks can also be acquired on Java functions which are imported and invoked from an XQuery expression. It is advisable to explicitly lock Java code whenever it performs sensitive read and write operations.

Limitations

Commands

Database locking works with all commands unless the glob syntax is used, such as in the following command call:

  • DROP DB new*: drop all databases starting with "new"

XQuery

Deciding which databases will be accessed by a complex XQuery expression is a non-trivial task. Database detection works for the following types of queries:

  • //item, read-locking of the database opened by a client
  • doc('factbook'), read-locking of "factbook"
  • collection('db/path/to/docs'), read-locking of "db"
  • fn:sum(1 to 100), locking nothing at all
  • delete nodes doc('test')//*[string-length(local-name(.)) > 5], write-locking of "test"

All databases will be locked by queries of the following kind:

  • for $db in ('db1', 'db2') return doc($db)
  • doc(doc('test')/reference/text())
  • let $db := 'test' return insert nodes <test/> into doc($db)

You can consult the query info output (which you find in the Info View of the GUI or which you can turn on by setting QUERYINFO to true) to find out which databases have been locked by a query.

File-System Locks

Update Operations

During a database update, a locking file upd.basex will reside in that database directory. If the update fails for some unexpected reason, or if the process is killed ungracefully, this file will not be deleted. In this case, the database cannot be opened anymore, and the message "Database ... is being updated, or update was not completed" will be shown instead.

If the locking file is manually removed, you may be able to reopen the database, but you should be aware that database may have got corrupt due to the interrupted update process, and you should revert to the most recent database backup.

Database Locks

To avoid database corruptions that are caused by accidental write operations from different JVMs, a shared lock is requested on the database table file (tbl.basex) whenever a database is opened. If an update operation is triggered, and if no exclusive lock can be acquired, it will be rejected with the message "Database ... is currently opened by another process.".

Please note that you cannot 100% rely on this mechanism, as it is not possible to synchronize operations across different JVMs. You will be safe when using the client/server or HTTP architecture.

Changelog

Version 8.6
  • Updated: New FAIRLOCK option, improved detection of lock patterns.
Version 7.8
Version 7.6
  • Added: database locking introduced, replacing process locking.
Version 7.2.1
  • Updated: pin files replaced with shared/exclusive filesystem locking.
Version 7.2
  • Added: pin files to mark open databases.
Version 7.1
  • Added: update lock files.