30 Jul, 2010

3 commits

  • git-svn-id: https://svn.forgerock.org/opendj/trunk@6558 41b1ffd8-f28e-4786-ab96-9950f0a78031
    ludovicp
     
  • There are 2 components:
    Index Filter Analyzer : Part of DatabaseEnvironmentProvider, it gathers search filter statistics and displays the index filter, number of hits, max matching entries, and a message. Compound filters are broken down to their basic elements. f an index was not utilized while evaluating the search filter, max matching entries will be -1 and a diagnostic message will be included. The monitor entry also includes when the analyzer was enabled along with the number of total index and unindexed searches processed. 
    
    Attribute Index Analysis, part of the dbtest tool. It cursors through all the indexes and finds the number of 100%, 95%, 90%, and 80% undefined index keys.  It also displays the value of the undefined keys along with the total number of keys in the index. The process is currently single threaded and might not scale to 10 mil+ DBs. However, this is currently unavoidable without the ability for JE to cursor through the records in log order.
    
    
    git-svn-id: https://svn.forgerock.org/opendj/trunk@6555 41b1ffd8-f28e-4786-ab96-9950f0a78031
    ludovicp
     
  • …or exiting of disks full.
    
    Each instance of the monitor have two thresholds: "low" and "full". Other components can register their own handlers if they wish to receive notifications when the thresholds are reached. It extends the MonitorProvider interface and utilizes the update interval mechanism to poll the free disk space. The current free space and state of all monitor instances are exposed through the cn=monitor interface under the cn=Disk Space Monitor branch. The only component using the monitor now is the JE backend. When the "low" threshold is reached, write operations will only be permitted to users with the BYPASS_LOCKDOWN privilege. When the "full" threshold is reached, the backend is placed in read-only mode. The default "low" and "full" thresholds are 100 MB and 10MB respectively. Along with the new feature, this patch also added two privileges: SERVER_LOCKDOWN and BYPASS_LOCKDOWN. Any user with the SERVER_LOCKDOWN privilege can put and take the server out of lockdown mode. Any user with the BYPASS_LOCKDOWN mode can bypass lockdown mode. This was restricted to only "root" users before. In addition, the MonitorProvider abstract class no longer extends DirectoryThread. It now uses a static single threaded ScheduledExecutorService to update provider states. Any provider that wishes to have its state updated periodically needs to register its updater runnable with the scheduleUpdate method.
    
    git-svn-id: https://svn.forgerock.org/opendj/trunk@6549 41b1ffd8-f28e-4786-ab96-9950f0a78031
    ludovicp
     

05 Jul, 2010

1 commit

  • The max log size is decreased to 10MB and checkpointer to 20MB.
    Also increases the number of cleaners to match the default number of workers.
    These new settings are helping with reducing the etimes on write operations
    
    git-svn-id: https://svn.forgerock.org/opendj/trunk@6537 41b1ffd8-f28e-4786-ab96-9950f0a78031
    ludovicp
     

25 Jun, 2010

1 commit


07 Jun, 2010

1 commit


31 May, 2010

2 commits


28 May, 2010

1 commit


27 May, 2010

1 commit


29 Apr, 2010

1 commit

  • This adds support for the IETF based Password Policy for LDAP as SubEntry.
    Also resolves the following issues :
    - 4544 :  initializeBackend() should not set JE env config params directly.
    - 4478 : ECL in draft compat mode / search lastchangenumber can be very long
    - 4538 : Virtual attributes not retrieved when entry cache configured
    - 4547 : Search Filter Matching differ for cn=Directory Manager and plain user.
    - 4514 : Logs shows unexpected message with replication monitoring data missing (Partial fix)
    - 4534 : Replication using security does not work after server restart
    - 4516 : SEVERE_ERROR: servers (...) have the same ServerId
    In addition, they also improve reliability and performance in various areas including CollectiveAttributes, Virtual Attributes and Subentries management, Schema loading, Replication...
    
    git-svn-id: https://svn.forgerock.org/opendj/trunk@6400 41b1ffd8-f28e-4786-ab96-9950f0a78031
    ludovicp
     

27 Apr, 2010

1 commit

  • Also improves unit, functional tests.
    More specifically this commit resolves the following open issues:
    4385 - NPE when using ExtensibleMatch filter without a matching rule
    4521 - dynamic lookup in attribut selection when selecting the sort order attribut while defining VLV index
    4531 - Control Panel creates virtual static groups using groupOfURLs as objectclass
    4533 - NullPointerException when configuring replication between 2 OpenDS
    4539 - DSML Gateway - jaxb.properties Exception
    
    
    git-svn-id: https://svn.forgerock.org/opendj/trunk@6396 41b1ffd8-f28e-4786-ab96-9950f0a78031
    ludovicp
     

19 Jan, 2010

2 commits

  • …0MBto 100MB): decrease default checkpoint frequency from 20MB to 100MB in order to reduce frequency of fsyncs. This should not increase recovery times significantly since modern hardware is capable of recovering 100MB of log very quickly.
    
    
    
    git-svn-id: https://svn.forgerock.org/opendj/trunk@6364 41b1ffd8-f28e-4786-ab96-9950f0a78031
    matthew_swift
     
  • replication server it will connect to. This also introduces a new 
    (weighed) load balancing feature that spreads DS connections across the 
    RSs, according to the RS weights defined by the administrator,
    
    Issue 4343: https://opends.dev.java.net/issues/show_bug.cgi?id=4343
    
    The commit includes necessary modifications for implementing what is 
    described in this document:
    
    https://www.opends.org/wiki/page/ReplicationServerSelection
    
    There is a little "implementation" section in this document that 
    explains some important stuff.
    
    Also good to know for these code modifications:
    
    - The topology info related to RSs and kept by the ReplicationBroker is 
    now fully kept in a new map of a new ReplicationServerInfo bag class. 
    This map is updated upon reception of a TopologyMsg.
    - Protocol change: the TopologyMsg now includes the RS url in the RSInfo 
    list
    - The dynamic change of the weight of a RS triggers a new TopologyMsg 
    being fired, to support dynamic change of weights and automatic topology 
    re-connections
    - SameGroupIdPoller thread has disappeared and its functionality is 
    replaced by the mechanism that re-evaluates the more suitable RS (see 
    section 5, in the document for more details)
    
    git-svn-id: https://svn.forgerock.org/opendj/trunk@6362 41b1ffd8-f28e-4786-ab96-9950f0a78031
    mrossign
     

14 Jan, 2010

1 commit


12 Jan, 2010

1 commit


25 Dec, 2009

1 commit


10 Nov, 2009

2 commits


06 Nov, 2009

2 commits

  • - Allow 0 value for monitoring publisher period to allow disable it by configuration 
    
    
    
    git-svn-id: https://svn.forgerock.org/opendj/trunk@6105 41b1ffd8-f28e-4786-ab96-9950f0a78031
    mrossign
     
  • we introduce:
    
    
    
    - a weigth, which is an integer affected to each RS that combined with 
    each others will define a percentage value which matches the number of 
    DSs (compared with total number od DSs in the topology) that can be 
    connected to the RS at a time in the topology. In these modif, this 
    configuration of the weight is added as well as dynamic changes. Also 
    transported in Topo messages. No modification of the connection 
    algorithm yet
    
    
    
    - Also to support the future connection algorithm, these modifs 
    introduces a Monitoring Publisher thread which is a thread that sens 
    every 3 seconds a Monitoring message (unchanged format) to every DSs 
    that are connected to him. These information will be used by the DSs to 
    potentially reconnect to another RSs with a newer server state (info 
    included in monitoring messages)
    
    
    
    The new connection algorithm will take into account:
    
    - group id
    
    - generation id
    
    - server states
    
    - locality (same VM)
    
    - weight (load)
    
    
    
    git-svn-id: https://svn.forgerock.org/opendj/trunk@6097 41b1ffd8-f28e-4786-ab96-9950f0a78031
    mrossign
     

20 Oct, 2009

1 commit


09 Oct, 2009

3 commits


08 Oct, 2009

1 commit

  • in handshake phase instead of a ReplServerStartMsg. ReplServerStartDSMsg 
    contains same thing as ReplServerStartMsg but also contains
    
      - replication server weight
    
      - number of ciurrently connected DS on the RS
    
      => both will be used for future new RS choice algorithm
    
    
    
    - Addition of a StopMsg sent:
    
      - when any entity (DS,RS) is closing a connection (sent just before) 
    with a peer
    
      - when DS finishes phase 1 of handshake (was gathering RS info for RS 
    choice so sent just after new ReplServerStartDSMsg is received)
    
      => both are used to distinguish between a proper connection closure 
    (no message) and an unexpected one (error log)
    
    
    
    - Compatibility between protocol V4 and V3 (and before)
    
      - changed MonitorMsg to never be created with a protocol version
    
      - MonitorMsg now always sent with publish(msg, version) (publish 
    method without version was used so bug)
    
      - TopologyMsg now always sent with publish(msg, version) (publish 
    method without version was used so bug)
    
    
    
    git-svn-id: https://svn.forgerock.org/opendj/trunk@5950 41b1ffd8-f28e-4786-ab96-9950f0a78031
    mrossign
     

07 Oct, 2009

1 commit


06 Oct, 2009

1 commit


29 Sep, 2009

1 commit


22 Sep, 2009

1 commit


16 Sep, 2009

1 commit


14 Sep, 2009

1 commit


07 Sep, 2009

1 commit


26 Aug, 2009

2 commits


20 Aug, 2009

1 commit


17 Aug, 2009

1 commit


14 Aug, 2009

1 commit


10 Aug, 2009

1 commit


05 Aug, 2009

1 commit

  • 4161 some mistranslations in german
    and
    4162 french: exception in setup
    
    
    git-svn-id: https://svn.forgerock.org/opendj/trunk@5631 41b1ffd8-f28e-4786-ab96-9950f0a78031
    hajma