« Changelog History


ArangoDB v1.4.0-beta2 Release Notes

Release Date: 2013-10-14 // almost 12 years ago
    • 🛠 fixed compaction on Windows

    The compaction on Windows did not ftruncate the cleaned datafiles to a smaller size. This has been fixed so not only the content of the files is cleaned but also files are re-created with potentially smaller sizes.

    • only the following system collections will be excluded from replication from now on:
      • _replication
      • _trx
      • _users
      • _aal
      • _fishbowl
      • _modules
      • _routing

    Especially the following system collections will now be included in replication:

    • _aqlfunctions
    • _graphs

    In previous versions of ArangoDB, all system collections were excluded from the replication.

    The change also caused a change in the replication logger and applier: in previous versions of ArangoDB, only a collection's id was logged for an operation. This has not caused problems for non-system collections but for system collections there ids might differ. In addition to a collection id ArangoDB will now also log the name of a collection for each replication event.

    The replication applier will now look for the collection name attribute in logged events preferably.

    • ➕ added database selection to arango-dfdb

    • 🏁 provide foxx-manager, arangodump, and arangorestore in Windows build

    • ArangoDB 1.4 will refuse to start if option --javascript.app-path is not set.

    • ➕ added startup option --server.allow-method-override

    This option can be set to allow overriding the HTTP request method in a request using one of the following custom headers:

    • x-http-method-override
    • x-http-method
    • x-method-override

    This allows bypassing proxies and tools that would otherwise just let certain types of requests pass. Enabling this option may impose a security risk, so it should only be used in very controlled environments.

    The default value for this option is false (no method overriding allowed).

    • ➕ added "details" URL parameter for bulk import API

    Setting the details URL parameter to true in a call to POST /_api/import will make the import return details about non-imported documents in the details attribute. If details is false or omitted, no details attribute will be present in the response. This is the same behavior that previous ArangoDB versions exposed.

    • ➕ added "complete" option for bulk import API

    Setting the complete URL parameter to true in a call to POST /_api/import will make the import completely fail if at least one of documents cannot be imported successfully.

    It defaults to false, which will make ArangoDB continue importing the other documents from the import even if some documents cannot be imported. This is the same behavior that previous ArangoDB versions exposed.

    • ➕ added missing swagger documentation for /_api/log

    • 🔊 calling /_api/logs (or /_admin/logs) is only permitted from the _system database now.

    Calling this API method for/from other database will result in an HTTP 400.

    ' ported fix from https://github.com/novus/nvd3/commit/0894152def263b8dee60192f75f66700cea532cc

    This prevents JavaScript errors from occurring in Chrome when in the admin interface, section "Dashboard".

    • 🌐 show current database name in web interface (bottom right corner)

    • ➕ added missing documentation for /_api/import in swagger API docs

    • 👍 allow specification of database name for replication sync command replication applier

    This allows syncing from a master database with a different name than the slave database.

    • issue #601: Show DB in prompt

    arangosh now displays the database name as part of the prompt by default.

    Can change the prompt by using the --prompt option, e.g.

      > arangosh --prompt "my db is named \"%d\"> "