Vendor and configuration parameters (Centera)

This help topic describes Centera configuration parameters.

Type
The list displays those fixed content device providers certified for the present release.
Vendor
This read-only field displays the fixed content device provider vendor's name.
Test Connection
When you click the Test Connection button, Enterprise Manager tries to log onto the fixed content product using the parameters entered on this page (described below). It then displays a success or failure message.

Configuration Parameters (Centera)

Attribute Name Attribute Value
FCP Pool Idle Timeout Seconds

Integer
Default=600
The number of seconds an idle connection can remain in the pool before being eligible to be closed and removed from the pool. The value must be greater than or equal to zero. Values less than zero treated as zero.

FCP Pool Preferred Size

Integer
Default=5
The preferred size of provider pool. Values less than zero treated as zero.

FCP Pool Max Wait Seconds Integer
Default=5
The maximum number of seconds to wait for a connection once the FcpPoolMaxInUse limit has been reached.
FCP Pool Max In Use

Integer
Default=-1 (unlimited)
The maximum number of concurrent connections to the fixed content device.

Pool Address

String
The Centera Pool Address, passed to FPPool_Open as the 'PoolAddress' parameter. Enter the DNS name, or IP address of the fixed content device, expressed in dotted quad format (decimal values of its 4 bytes separated by periods). For example, 127.0.0.1. If more than one access node is used, separate the entries with a comma. This parameter also supports the use of a PEA file. For example:

10.10.16.1,10.10.16.2?c:\centera\my.pea

Or, if the file is shared by more than 1 computer, the filename is a shared path name:

10.10.16.1,10.10.16.2?\\myhost\centera\my.pea

NOTE   For additional information on available connection string options, refer to the EMC Centera documentation.

An advanced technique is to write the XML (see About Centera > Pool Address Site Mapping XML Data for an example) to a UNICODE file (prefix the XML with "sitemap"), then set the Centera connection string to the file name (with ‘file:’ prefix). Consider placing the file in the P8 domain GCD share, since it's already shared and accessible from all Content Engine service computers in the P8 domain, has proper security settings, and should be part of regular system backups.

NOTE   If the Centera fixed content device has a sitemap file configured, make sure that it is encoded as a UTF-16 file. Otherwise, any attempt to migrate content to the Centera device will fail.

For example, if the file ‘Map_Centera.xml’ is copied to the GCD share, set the connection string to ‘file:\\server\FNCE_server$\Map_Centera.xml’.

Concurrent Readers

Integer
Used to limit the maximum number of concurrent Centera reader threads. A semaphore is used to insure that a given OSS process will only allow a maximum of Concurrent Readers to simultaneously call FPTag_BlobReadPartial.

Max Reader Semaphore Wait Time The maximum waiting time to get a read semaphore for read operation before the system fails the operation.
Default Retention Pass Through

Boolean
This flag only applies when a fixed storage area's Default Retention Period is configured to use the Same as the retention period of the fixed content device. Normally the provider will examine the value of the FP_RETENTION pool capability and explicitly set the retention period to either zero or infinite, depending on the capability setting. When Default Retention Pass Through is set to 1 (TRUE), the retention period is set to FN_RETENTION_PERIOD_DEFAULT (meaning that the Clips use the current Centera system setting, which can be altered by changing the compliance mode of the system).

Never Delete Clips Or Content

Boolean
If set to 1 (TRUE), the Centera provider never attempts to delete Clips or BLObs, all content is left completely intact on the Centera device. Note that normally this setting has no impact on enforcement of the retention period - Content Engine still honors the retention period and prevents delete when appropriate. This flag is only examined by the low level delete routine within the provider; if TRUE, the delete is not processed.

Note that referrals can still be deleted from the store (as well as content for reservations and annotations).

Set this flag to 1 if you want to run Centera in a write-once amd delete-never mode, but also want to be able to delete documents from the object store.

Blob Write Collision Avoidance Flag Integer
Used by the provider when creating Centera content, the value is combined with FP_OPTION_CLIENT_CALCID_STREAMING to form the options parameter of FPTag_BlobWrite. There are three possible [decimal] values:
  • 0 - Use the collision avoidance scheme of the Centera system.
  • 64 - FP_OPTION_ENABLE_COLLISION_AVOIDANCE override system setting, enable collision avoidance.
  • 128 - FP_OPTION_DISABLE_COLLISION_AVOIDANCE override system setting, disable collision avoidance
Blob Read Ahead Size Integer
This parameter supplies a maximum buffer size, used by the provider to determine if the entire content blob should be read with a single call to FPTag_BlobRead. If the content size is less than or equal to the parameter value, the entire blob is read using a single call (a buffer of content size must be allocated), and the data is passed back via the callers 'stream' as requested. If the content size is greater than the parameter value, FPTag_BlobReadPartial is used to read only the data of the size requested by the caller (no buffer is created).
Embedded Data Threshold

Integer
This parameter supplies the maximum data size per content element to embed in the C-Clip. By default this value is zero, meaning that content is never embedded in the C-Clip. The maximum value for this parameter is 102400 (bytes), if a greater value is supplied it will be reduced to 102400 internally by the Centera provider.

Note that this limit is per content element. The overall size of embedded data can exceed this limit if multiple content elements are embedded in a single Clip (up to the maximum size of the C-Clip).

Audited Delete Prefix String
This parameter supplies a prefix string to use with Centera audited delete. If the string is blank, audited delete will not be invoked (FPClip_Delete will be used). If the string is not blank, then the string will be used as the prefix of the delete reason passed to FPClip_AuditedDelete (the document ID GUID is appended to the prefix string to form the complete delete reason text).
Buffer Size

Integer
The size of an internal C-Clip buffer in bytes. The default value is 16*1024. The minimum value is 1 KB. The maximum value is 10 MB.

Timeout Integer
TCP/IP connection timeout in milliseconds. The default value is 120000 ms (2 minutes). The maximum value is 600000 ms (10 minutes)
Enable Multicluster Failover Boolean
When this option is true (the default), multicluster failover is enabled. By default this option is true (1). To turn multicluster failover off for all capabilities, specify false (0).
Default Collision Avoidance Binary
"This option can either be true (1) or false (0). This option is false by default. To enable collision avoidance at pool level set this option to true. If you enable this option, the SDK uses an additional blob discriminator for read and write operations of C-Clips and
blobs. (Blob Write Collision Avoidance Flag is used in tag write and only applies to blobs)"
Prefetch Size Integer
The size of the prefetch buffer. This buffer is used to assist in determining the size of the blob. The default size is 32 KB. The maximum size is 1 MB.
Maximum Connections Integer
The maximum number of sockets that the SDK will allocate for your application. Sockets are used to communicate with the Centera clusters managed in each pool object. The default value is 100. The maximum value is 999.
Retry Count Integer
The number of times an operation will be retried before a failure is reported to the client
application. The default value is 6. If the first execution of the function fails, the system retries the function 6 times. In total the function executes 7 times. The maximum value is 99.
Retry Sleep Integer
The time to wait before the failed API function call should be retried, in milliseconds. The maximum value is 100000 ms. If no retry sleep has been defined, the SDK uses an exponential back-off scheme. The sleep time increases after each retry, starting at 1 second, and doubles after each retry
Cluster Nonavailable Time Integer
The time in seconds that a cluster is marked as not available before retrying with a probe. Other clusters in the pool will be used while the cluster is unavailable. The default value is 600 (10 minutes). The minimum is 0. The maximum is 36000 (10 hours).
Open Strategy

Integer
The approach used by FPPool.Open() to open connections to addresses in the connection string. Choices are:

  • FP_NORMAL_OPEN(0) — FPPool_Open() attempts to open
    connections to all addresses in the connection string, and to
    all associated replication addresses. Consider using this
    strategy if your application performs numerous operations
    while the pool is open. This strategy is the default. This
    option is equivalent to FP_OPTION_DEFAULT_OPTIONS.
  • FP_LAZY_OPEN (1)— FPPool_Open() opens connections to
    addresses only as needed. Consider using this strategy if
    your application frequently opens and closes the pool.
Multicluster Read Strategy

Integer
The multicluster failover strategy for read operations: FPClip_Open(), FPTag_BlobRead(), FPTag_BlobReadPartial().

Choices are:

FP_NO_STRATEGY(0)
FP_FAILOVER_STRATEGY(1)
FP_REPLICATION_STRATEGY (2, default)

Multicluster Write Strategy

Integer
The multicluster failover strategy for write operations: FPClip_Write(), FPTag_BlobWrite().

Choices are:

FP_NO_STRATEGY(0, default)
FP_FAILOVER_STRATEGY(1, Not supported)
FP_REPLICATION_STRATEGY (2)

Multicluster Delete Strategy

Integer
The multicluster failover strategy for delete operations:

FPClip_Delete(), FPClip_AuditedDelete(). This option also controls failover for the deprecated purge operations: FPClip_Purge(), FPTag_BlobPurge().

Choices are:

FP_NO_STRATEGY(0, default)
FP_FAILOVER_STRATEGY(1, Not supported)
FP_REPLICATION_STRATEGY (2)

Multicluster Exists Strategy

Integer
The multicluster failover strategy for exists operations: FPClip_Exists(), FPTag_BlobExists().

Choices are:

FP_NO_STRATEGY(0)
FP_FAILOVER_STRATEGY(1, Default)
FP_REPLICATION_STRATEGY (2)

Multicluster Query Strategy

Integer
The multicluster failover strategy for query operations: FPPoolQuery_Open().

Choices are:

FP_NO_STRATEGY(0)
FP_FAILOVER_STRATEGY(1, Default)
FP_REPLICATION_STRATEGY (2)