Last revision: 10/7/2008
This Readme file describes the defects fixed in this Wonderware Historian patch and the known issues. The Wonderware Historian was formerly called IndustrialSQL Server™.
Click on the following links to go to the various sections of this Readme.
Release Summary - Wonderware Historian 9.0 Patch 02
System Requirements - Wonderware Historian 9.0 Patch 02
Installation Instructions - Wonderware Historian 9.0 Patch 02
Resolved Issues - Wonderware Historian 9.0 Patch 02
Known Issues - Wonderware Historian 9.0 Patch 02
Installation Instructions - Wonderware Historian 9.0 Patch 01
Resolved Issues - Wonderware Historian 9.0 Patch 01
Known Issues - Wonderware Historian 9.0 Patch 01
Wonderware Historian 9.0 Features
Known Issues – Wonderware Historian
Known Issues - Wonderware Historian Running in a Cluster
Wonderware Historian Documentation Issues
Wonderware Historian Documentation
This patch includes the following:
If you are using Windows 2000 and Microsoft SQL Server 2000, you must install .NET Framework 2.0.
The Wonderware Historian 9.0 Patch 02 is supported on the following operating systems and versions of Microsoft SQL Server. For detailed system requirements, see the IndustrialSQL Server Historian Installation Guide (InSQLInstall.pdf) or installation Help.
Wonderware Historian:
Wonderware Historian Configuration Tools and remote IDAS:
We strongly recommend that you exit all Wonderware® programs, including executable (.exe) files and services, before you install this patch.
The Wonderware Historian needs to be shut down and disabled to install this patch.
Note:
If you want to avoid a reboot on completing the
installation, stop, and restart the Microsoft SQL Server. This unloads the InSQL
OLE-DB provider.
The patch should be installed on all
Wonderware Historian
component computers, including servers, remote IDASs, and remote client consoles.
Before you install Historian 9.0 Patch 02 on a clustered SQL Server 2005 system, or a non-clustered SQL Server 2005 Enterprise Edition system, make sure that you have the proper license file for operation on your system. For more information, see CRs L00080162 and L00080164. These CRs describe license enforcement issues addressed in this patch. Wonderware may have provided you a workaround to these problems in Historian 9.0 or 9.0 Patch 01 by delivering a replacement license file. If you use such a replacement license file with Patch 02, the Historian license file check fails and the system stops storing data.
Using the Cluster Administrator, take the Wonderware Historian offline. (This will Shutdown and Disable the Wonderware Historian).
During the patch install, you will be required to enter Cluster configuration information. It is important to obtain current accurate information. This information can be obtained from the Cluster Administrator. In the Cluster Administrator, under Groups, select the Disk Group containing the following resources:
Historian Network Name:
The name for the virtual Wonderware Historian should appear in
parenthesis beside the Name for this resource. You can also right-click on this
resource and select ‘Properties’. On the Properties dialog, select the
‘Parameters’ tab and note the ‘Name’ parameter.
Historian IP Addresses:
For each Historian IP address resource, right-click on the resource and
select ‘Properties’. On the ‘Properties’ dialog, select the ‘Parameters’ tab and
note the IP address, Subnet mask, and Network.
Also, the following information will need to be obtained from the Runtime database either via querying the database, or by looking at the Wonderware Historian configuration editor. The information is found under System Configuration/Storage/Storage Locations.
Wonderware Historian Data
Paths:
In Wonderware Historian, note the current paths where the Wonderware Historian
data is stored (for example: Y:\InSQL\DATA).
On the active node, run the patch install. Move the Disk Group containing the Historian resources to the second node. Run the patch install on the second node.
After the installation is complete, verify that the
Wonderware Historian data paths are correct.
The following defects were fixed in this patch release.
| L00045575: | The buffer count used by MDAS is configurable using a Registry setting. | ||
| L00068433: | If you shut down the Historian and then immediately reboot the computer, frequent disconnects/reconnects occur for MDAS clients, such as the Wonderware Application Server.This fix requires Application Server 3.1. | ||
| L00063856: | If you use both >= and <= in the DateTime criteria and wwCycleCount=1, the results have one fewer row than expected. The query should return the same amount of rows for each tag without excluding any of the tags from the list. | ||
| L00075551: | Performing a minimum retrieval returns an initial value row with a QualityDetail value of 133, if you use the < and > signs instead of <= and >= signs. | ||
| L00075553: | Peforming a minimum, maximum, and "best fit" retrieval does not return all expected data values, if you use the < and > signs instead of <= and >= signs. | ||
| L00077833: | During tag registration, the Wonderware Historian discards the value for the Sample Count option that is configurable in Application Server. | ||
| L00078025: | The "swinging door" storage filter should be applied to data coming from ArchestrA. | ||
| L00079301: | When swinging door is not enabled and the object is redeployed, the Wonderware Historian stops storing data values. | ||
| L00079303: | History block changeover fails during on daylight savings time change with HoursPerBlock=2 and HoursPerBlock=1. | ||
| L00079971: | Calling the Replace() method twice in a quick sequence for the
same tag with two different time ranges deletes the data between the
replaced time ranges. | ||
| L00079988: | A delta or full re-import does not import changes to a tag comment from the InTouch HMI. | ||
| L00079989: | The OpenQuery syntax leaks memory of the Wonderware aahRetSvc service. | ||
| L00079990: | If you register the Wonderware Historian from a remote machine and then log in as a domain user who is not a member of the Administrators Windows group on the Wonderware Historian, you get the error: "New server name | ||
| L00079992: | After an AppServer engine failover, a NULL value with a QualityDetail of 10 is stored for slow-changing tags. No data value is received from the latest active engine. Thus, a data hole is created until the value changes. | ||
| L00079993: | Counter retrieval mode returns zero when the last value does not change. | ||
| L00079994: | Counter retrieval mode returns a NULL value if there is bad data at the end of the query period. | ||
| L00079995: | The counter retrieval mode detects a false rollover when bad data is present at the beginning of the period, and thus returns a counter value that is too large. | ||
| L00079996: | Insert operations using normal CSV files can cause erroneously written snapshot data to prevent correct retrieval of original values. | ||
| L00080162: | License violation occurs for the Historian Clustered Enterprise Edition if installed on a Windows Server 2003 standard edition. The system incorrectly expects a "Historian_ServerOS" feature line instead of a "Historian_MSCluster" feature line. | ||
| L00080164: | License violation occurs with non-clustered system if installed on SQL Server 2005 Enterprise Edition. The system incorrectly expects a "Historian_MSCluster" feature line instead of a "Historian_ServerOS" feature line. | ||
| L00080225: | For the "best fit" retrieval mode, if a NULL is encountered in the source data, no data after the NULL is returned. | ||
| L00080346: | A query that uses wwCycleCount=1 AND wwRetrievalMode='Cyclic' returns five rows for three tags. | ||
| L00083028: | In certain cases, and during repeated network disconnects/reconnects, manual storage crashes when store-and-forward blocks are merged, caused by memory corruption. | ||
| L00083030: | After a fail-over for a cluster system, MDAS does not detect the Historian is back up and running. | ||
| L00084589: | A warning message is logged for each tag when MDAS receives late data with OPC quality of INITIALIZING. |
The following defects are known issues in this patch release.
| L00068433: | If you shut down the Historian and then immediately reboot the computer, frequent disconnects/reconnects occur for MDAS clients, such as the Wonderware Application Server. Also, if you patch the Historian with Patch 02, the frequent disconnect/reconnect events will occur for any active Application Server 2.1 or 3.0 MDAS client. If you are using Application Server 2.1 or 3.0, you can stop the disconnect/reconnect cycle by redeploying the Galaxy. To completely fix this issue, you must install Application Server 3.1 and Historian 9.0 Patch 02. | ||
| L00084146: | For any node in a cluster configuration, if you install Historian 9.0 Patch 02 on a Historian 9.0 installation that has been taken offline, you may receive the following error: "Failed to create Resource: 'IndustrialSQL Server'. Error Code: 00001392. Error Message: The object already exists." If you install the Historian 9.0 and then immediately install Patch 02 without starting up the Historian, you do not get the error. Workaround: Before installing Historian 9.0 Patch 02, verify that the following registry keys are in place. In this example, "QAVINSQL" is the cluster name. Your own setup will be different. [HKEY_LOCAL_MACHINE\SOFTWARE\ArchestrA\Historian\Setup\Cluster] "ClusterName"="QAVINSQL" [HKEY_LOCAL_MACHINE\SOFTWARE\ArchestrA\Historian\Setup\Cluster\Resources] "NetworkName"="Historian Network Name (QAVINSQL)" "IP Address1"="Historian IP Address 1 (QAVINSQL)" "IP Address2"="Historian IP Address 2 (QAVINSQL)" "ResourceName"="IndustrialSQL Server" | ||
| L00084218: | The PublicNameSpace table in the Runtime database may grow extremely large if you deploy Application Server objects that are historizing to the Wonderware Historian. Under certain conditions, the deployment of historized Application Server objects may unexpectedly generate multiple redundant namespace hierarchies in the Runtime database. This may result in excessive database size, increased CPU usage, and/or slow performance of the ActiveFactory TagPicker when trying to open or expand these redundant namespaces. A possible workaround for this issue can be found in Wonderware Tech Note #477. |
Last revision: 2/22/2007
This Readme file describes the defects fixed in this patch and known issues.
We strongly recommend that you exit all Wonderware® programs, including executable (.exe) files and services, before you install this patch.
The Wonderware Historian needs to be shut down and disabled to install this patch.
Specific installation procedure for Installing Patch 01 on a CLUSTERED Wonderware Historian 9.0
Using the Cluster Administrator, take the Wonderware Historian offline. (This will Shutdown and Disable the Wonderware Historian).
During the patch install, you will be required to enter Cluster configuration information. It is important to obtain current accurate information. This information can be obtained from the Cluster Administrator. In the Cluster Administrator, under Groups, select the Disk Group containing the following resources:
Historian Network Name:
The name for the virtual Wonderware Historian should appear in
parenthesis beside the Name for this resource. You can also right-click on this
resource and select ‘Properties’. On the Properties dialog, select the
‘Parameters’ tab and note the ‘Name’ parameter.
Historian IP Addresses:
For each Historian IP address resource, right-click on the resource and
select ‘Properties’. On the ‘Properties’ dialog, select the ‘Parameters’ tab and
note the IP address, Subnet mask, and Network.
Also, the following information will need to be obtained from the Runtime database either via querying the database, or by looking at the Wonderware Historian configuration editor. The information is found under System Configuration/Storage/Storage Locations.
Wonderware Historian Data
Paths:
In Wonderware Historian, note the current paths where the Wonderware Historian
data is stored (for example: Y:\InSQL\DATA).
On the active node, run the patch install. Move the Disk Group containing the Historian resources to the second node. Run the patch install on the second node.
After the installation is complete, verify that the Wonderware Historian data paths are correct.
Note:
If you want to avoid a reboot on completing the
installation, stop, and restart the Microsoft SQL Server. This unloads the InSQL
OLE-DB provider.
The patch should be installed on all Wonderware Historian component machines, including servers, remote IDAS, and remote client consoles.
In addition to the supported operating system and Microsoft SQL Server versions listed in the System Requirements of the 9.0 release, this patch also supports Microsoft SQL Server 2005 (at SP1 level), Standard and Enterprise editions.
The following defects were fixed in this patch release.
| L00040869 : | If the Index service consumes memory up to the limit of its processing space (normally 2GB), it will crash and cause a SmartHeap error.
The effective HistoryCacheSize is always limited by the following parameters regardless of the value that is specified by the user in the System Parameters table:
By default when the HistoryCacheSize is 0, the
indexing service automatically uses its own HistoryCacheSize as 20% of
physical RAM, then applies the rules 1-3 mentioned above. |
||
| L00040870: | Indexing service should support large cache sizes on Windows Server 2003 Enterprise Edition.
The indexing service is limited to 2GB of process memory, even on system that can support up to 3GB. Note:If a /3GB boot.ini switch is specified for the operating system, the Indexing service can use more than 2GB memory for tag information if such HistoryCacheSize is specified above the 2GB limit. However, you should also take into account that the machine should have at least 4GB physical RAM installed and the retrieval performance may suffer because less memory is available to the OS for disk I/O buffering. |
||
| L00046335: | When Wonderware Historian 9.0 imports tags from the InTouch 9.5 HMI, the Tag Import wizard takes the secondary source as the configuration.
When Wonderware Historian 9.0 imports tags from the InTouch 9.5 HMI, the Tag Import wizard takes the secondary source as the configuration for IDAS topic, not the primary source as it should. |
||
| L00047220: | The Wonderware Historian 9.0 wide query does not return correct results. The next-to-last row is duplicated. | ||
| L00050134: | The Bestfit query returns no initial row in a specific scenario, causing incorrect plotting in the Trend.
The Bestfit query returns no initial row in a specific scenario. If the tag setting for interpolation type is set to 'Interpolated' or if the query overrides the interpolation type to force linear interpolation, the bestfit query will fail to return an initial row, if the point following the last point prior to the query start time is not located within the first cycle. A narrow query just fails to return the initial row, while a wide query returns a NULL at the query start time for the tag in question. This problem is particularly visible with slow changing rate tags. |
||
| L00051447: | wwRetrievalMode = ‘ValueState’ does not include in the StateTime calculation if the last sample is NULL.
In a query with wwRetrievalMode = 'ValueState' and wwStateCalc = 'Percent', if the last value of the sample is Null, then it is not included in the calculation. |
||
| L00057312: | From same tagname and time stamp, InSQLIOS has good OPCQuality - Wonderware Historian Storage and Retrieval shows bad OPCQuality.
By default, the Wonderware Historian I/O Server sends the QualityDetail as the quality of the point. However, if you want to send the original OPC quality stored in the Wonderware Historian, use the following workaround:
After this workaround is in place, Wonderware Historian I/O Server sends the original OPC quality as it was stored in Wonderware Historian, and not as in the QualityDetail. |
||
| L00059105: | A slope query with a start time in the future causes the retrieval service to shutdown and restart.
A required NULL row was missing when the start time was in the future. |
||
| L00060616: | Public Name Spaces are corrupted after Tags are deleted using the Wonderware Historian Configuration Editor and "orphan" records are left in the aaAttributeData table.
A redeploy after corruption of the PublicNameSpace did not complete. Tags would be mapped to the "All xxxTags..." namespace but not under the proper area. Therefore, Wonderware Historian did not reflect changes to the galaxy. |
||
| L00061498: | Rename of the Contained name of the objects are not updated in Wonderware Historian after deployment of the object which leads to wrong view in ActiveFactory.
The ActiveFactory view was not in sync with the actual model view of the deployed galaxy because Wonderware Historian failed to update the PublicNameSpace table correctly after receiving a model view from Wonderware Application Server. |
||
| L00061499: | Use of wwRetrievalMode = Counter requires a zero numeric value to be present.
The wwRetrievalMode counter should return a consistent value. A zero (value) should be stored in connection with a rollover selection. |
||
| L00061501: | High CPU load observed on Historian's SQL Server when deploying large AppServer application with more areas/objects.
The problem was due to stored procedures that are not optimized and an absence of indexes on the tables that deal with hierarchy namespace replications. In addition, the CPU load is high for a long time during failover when compared to the initial deployment because the stored procedure's queries have to go through an update sequence rather than just an insert. Note: |
Note:In addition to the known issues specifically mentioned in this section for Wonderware Historian 9.0 Patch 01, please also see the known issues for the initial release of Wonderware Historian 9.0.
| L00062377 : | OpenQuery AnalogWideHistory columns being
flagged as duplicates using SQLServer 2005 when using column names more than
once.
Workaround:When using a column name more than once in an OpenQuery syntax (for example, while performing calculations), perform the calculation outside the OpenQuery syntax (for example, on the result set) or make sure to use a column alias within the OpenQuery syntax. Examples:Fails: select * from OpenQuery(INSQL, 'select Column1, Column2, Column1 + Column2 from......') Works: select Column1, Column2, Column1 + Column2 from OpenQuery(INSQL, 'select Colum1, Column2 from.....') Works: select * from OpenQuery(INSQL, 'select Column1, Column2, MyCalculation = Column1 + Column2 from......') |
Last revision: 07/21/2005
This document contains important information about the
Wonderware® Wonderware Historian 9.0. Please review this
document for known issues in this version and changes to the Wonderware Historian documentation included on this CD.
New features of the Wonderware Historian 9.0 are as follows:
The Wonderware Historian 9.0 is supported on the following operating systems and versions of Microsoft SQL Server. For detailed system requirements, see the IndustrialSQL Server Historian Installation Guide (InSQLInstall.pdf) or installation Help.
Wonderware Historian:
Wonderware Historian Configuration Tools:
Remote IDAS:
The InSQLInstall.pdf file, located on the Wonderware Historian CD in the \UserDocs\English directory, includes comprehensive information on system requirements and recommendations, installation, upgrades and migration, and system sizing examples. If you are installing the Wonderware Historian or upgrading from a prior version, read this information prior to starting the installation or upgrade.
The CD-key for installing Microsoft SQL Server is: GJDD4-GMYXW-GQR4C-KT4P9-QDQGB
The InSQLEnterprise.pdf file, located on the Wonderware Historian CD in the \UserDocs\English directory, is a deployment guide that includes specific instructions about implementing the Wonderware Historian in a clustered environment.
The Microsoft SQL Server Enterprise Edition CD includes an embedded CD key. You are not be required to provide a CD key in order to install Microsoft SQL Server Enterprise Edition.
This section describes all known issues identified in the Wonderware Historian.
| Anti-Virus Software Issues : | Some virus scanners lock key files during a scan, and this may interfere with normal operation of the Wonderware Historian. Anti-virus software should not process files in the C:\InSQL folder. (This is the default installation folder for history data.)
If, during Wonderware Historian installation, you choose a different folder, be sure that you configure the anti-virus software accordingly. |
||
| L00011101: | When tags with the access name 'Galaxy' are imported into the Wonderware Historian, an I/O Server named "\NA\NA" is created, with a topic named "NA."
Workaround: The 'Galaxy' access name is from Wonderware Application Server tag definitions. InTouch applications that use Wonderware Application Server should not be imported into the Wonderware Historian directly from the Tagname.x file. Use the Configuration Export and Import Utility to remove the Galaxy items and then import the remaining tags into the Wonderware Historian. |
||
| L00011169: | Inconsistent date formats between the OPENQUERY function and four-part queries on non-English language systems.
The 'Galaxy' access name is from Wonderware Application Server tag definitions. InTouch applications that use Wonderware Application Server should not be imported into the Wonderware Historian directly from the Tagname.x file. Use the Configuration Export and Import Utility to remove the Galaxy items and then import the remaining tags into the Wonderware Historian. |
||
| L00012260: | SELECT INTO combined with ORDER BY fails.
With Microsoft SQL Server 2000 SP3, if you use a
SELECT...INTO statement from an Wonderware Historian extension table (e.g.
AnalogHistory) that includes an ORDER BY clause into a SQL Server table,
the statement will fail; without an ORDER BY, it will
succeed. "SELECT * INTO MyTable FROM AnalogHistory ORDER BY DateTime" and then: "SELECT * FROM MyTable" Instead, use: "SELECT * INTO MyTable FROM AnalogHistory" and then: "SELECT * FROM MyTable ORDER BY DateTime" | ||
| L00014736: | String tags are truncated to 511 characters for REALTIME and ORIGINAL inserts.
If you are inserting original or real-time data for string tags, the maximum length of the string value is 511 characters. If the string value is 512 characters, the 512th character is discarded. |
||
| L00014755: | Importing long string tag values from a CSV file.
When importing string data from a CSV file you
cannot combine strings longer than 256 characters into single line of the
CSV. |
||
| L00021328: | For the Wonderware Historian I/O Server (InSQLIOS), the item name is not case-sensitive.
If two tags have the exact same name but differ in case, the values for the tag that occurs first in the active image will be passed to Wonderware Historian I/O Server clients. |
||
| L00021564: | For single point updates via a CSV file, if the CSV file contains more than one value with the same ending timestamp for the same tag, the updates are not terminated (NULL value) until the next block change-over.
Workaround: Make sure that multiple updates in single CSV file do not have the same end time. |
||
| L00023669: | If the Wonderware Historian is running on the Windows Server 2003 operating system without Service Pack 1, the remote IDAS goes into store-and-forward mode at regular intervals, even though the connection to the Historian is still valid.
The Wonderware Historian running on Windows Server 2003 requires more processor time than the same configuration running under Windows 2000 Server, due to differences in the operating systems. For data storage tasks, expect Windows Server 2003 to require a processor approximately three times faster than a comparable system running Windows 2000 Server. While, in general, the Wonderware Historian quite effectively uses multiple processors, this particular limitation cannot be addressed with additional processors. For example, the dual 933MHz processor described in the system sizing examples is adequate for a 25,000-tag Wonderware Historian system running on Windows 2000 Server. Upgrading the system to Windows Server 2003 would require replacing the system with approximately a dual 2.7 GHz system rather than upgrading it to a quad 933MHz. These performance differences also applied to the Wonderware Historian 8.0 SP2, which was the first release to support Windows Server 2003. IDAS goes into store-and-forward mode because the storage subsystem does not process data sent by IDAS in timely fashion. In all cases, store-and-forward functions correctly, and data will not be lost. Workaround: Install Windows Server 2003 SP1. |
||
| L00028021: | If you use the Database Configuration Export and Import utility to import tag definitions with Japanese characters in the tag description, the error "Engineering unit does not exist" occurs.
Workaround: Convert the exported text file to SJIS, one of the Japanese character codes, before importing the file. |
||
| L00034386: | If dynamic reconfiguration occurs for a tag receiving late data, and a new history block is required, the late data is not available for retrieval from disk until the time span by which the data was late has expired.
For example, if the data is late by ten minutes, then data will be unavailable for that tag for ten minutes. |
||
| L00034465: | If you have upgraded to Wonderware Historian 9.0 from a previous version, you must change the SuiteLink client application name for the Wonderware Historian I/O Server from "InSQLIOS" to "aahIOSvrSvc." | ||
| L00037017: | If the Wonderware Historian installation does not require a computer restart, a "Stack overflow" exception in wwPerf.dll error occurs when the InSQLIOS service is started.
Workaround: Restart the computer after installing the Historian, even if not prompted to do so. |
||
| L00037116: | If the Wonderware Historian installation does not require a computer restart, IDASs fail to communicate to all local and remote I/O Servers because the Suitelink.dll could not be loaded.
Workaround: Restart the computer after installing the Historian, even if not prompted to do so. |
||
| L00037200: | The message "Failed to retrieve product name and/or version resources" is logged by the Wonderware Historian I/O Server and InSQLIOS (aahIOSvrSvc) during startup.
This message can be ignored. |
||
| L00037894: | After installing a valid license file, if you attempt to insert manual data using a SQL statement, Microsoft SQL Server returns an error.
Workaround: After installing a valid license file, completely shut down the Historian using the Shutdown (and disable) InSQL command and then restart the Historian using the Enable (allow to run) InSQL command. You can also restart the computer. If a running Historian cannot be interrupted, you can alternatively use a CSV file to insert the data until it is appropriate to take the Historian offline. |
||
| L00039362: | For "fast load" imports, the "swinging door" storage algorithm is not applied. | ||
| L00039808: | Warning message regarding snapshot write operation taking longer than 10 seconds needs to be more accurately informative.
This message means the data throughput rate of the system is exceeding the disk I/O write capabilities. If this situation persists, there is the possibility of data corruption and data loss. For example, this problem can occur on high throughput systems (updates/sec) and/or during extensive (circular path) disk activity generated by file copy operations. To solve this problem, reduce the data throughput
(for example, using deadbands) and/or to upgrade your hardware for optimal
disk access. |
This section describes all known issues identified in the Wonderware Historian Running in a Cluster.
| L00019269: | In some circumstances, when a cluster failover occurs due to a network failure on the active cluster node, NULLs will not be stored for system tags and I/O tags if the tag data is being acquired by a local IDAS. The resulting behavior is that after the system has failed over, there is a "flat line" during the failover time, rather than a gap. This expected behavior does not occur for remote IDAS data. |
||
| L00019335: | In some circumstances, if you shut down the Wonderware Historian on the active cluster node, a data loss of up to 30 seconds may occur before the store-and-forward functionality starts up on the remote IDAS. It is recommended that you move the cluster resources to the failover node prior to shutting down the Wonderware Historian. |
||
| L00019356: | In some circumstances, when a failover is initiated just after the reconnection of a remote IDAS (for example, from a previous network failure), the system may take up to five minutes to complete the failover. This does not affect the data collected by the remote IDAS, as this data will be stored and forwarded as soon as the failover is complete. |
||
| L00037750: | Uninstalling the Wonderware Historian from a cluster does not remove Historian resources and a resource type created during installation. Workaround: Run the following commands from a command prompt: REM delete Historian resource:cluster resource "IndustrialSQL Server" /delete REM delete Historian resource type:cluster restype Historian /delete /type REM take network name resource offline [name varies]cluster resource "Historian Network Name (< Virtual InSQL Server name here >)" /off REM delete network name resource [name varies]cluster resource "Historian Network Name (< Virtual InSQL Server name here >)" /delete REM take IP address resource offline [name varies]cluster resource "Historian IP Address 1 (< Virtual InSQL Server name here >)" /off REM delete IP address resource [name varies]cluster resource "Historian IP Address 1 (< Virtual InSQL Server name here >)" /delete REM take IP address resource offline [name varies]cluster resource "Historian IP Address 2 (< Virtual InSQL Server name here >)" /off REM delete IP address resource [name varies]cluster resource "Historian IP Address 2 (< Virtual InSQL Server name here >)" /delete |
The following table lists corrections to known issues in the Wonderware Historian documentation.
| Item | Location of Change | Existing Documentation | Change |
| 1 |
InSQLConcepts.pdf: "Swinging Door Deadband: Rate, Value, and Deadband Override Period" section, page 78 InSQLBooks.chm: "Swinging Door Deadband: Rate, Value, and Deadband Override Period" topic |
"To allow for better representation, a deadband override period may optionally be specified. If the elapsed time since the last stored point exceeds the deadband override period, the last received point before the time at which the deadband override period expired is stored, regardless of value and rate deadband. In this example, it would be point 9. The data actually stored to disk (points 1, 2, 9, and 15) is a better approximation of the original data." |
"To allow for better representation, a deadband override period may optionally be specified. If the elapsed time between the last stored point and the currently received point is more than the specified deadband, then the point immediately prior to the currently received point is stored. In this example, the elapsed time between point 2 and point 10 is more than the deadband, so point 9 is stored. The data actually stored to disk (points 1, 2, 9, and 15) is a better approximation of the original data." |
| 2 |
InSQLAdmin.pdf: "Formatting the CSV File for a Normal Import" section, page 144 InSQLBooks.chm: "Formatting the CSV File for a Normal Import" topic |
-- |
Add the following paragraph: "For a fast load CSV import, the end time of the current block in the block80.inf file is considered to be the current time, not the current system timestamp. The end time in the .inf file is updated by the storage subsystem every 20 seconds." |
| 3 |
InSQLConcepts.pdf: "Modification Tracking for Historical Data Changes" section, page 38 InSQLBooks.chm: "Modification Tracking for Historical Data Changes" topic |
"The count of the number of consecutive data value changes per tag is stored in the NewValue column of the ModLogColumn table." |
"The total count of consecutive value changes attempted per tag is stored in the NewValue column of the ModLogColumn table." |
| 4 |
InSQLAdmin.pdf: "Importing Data from an InTouch History File" section, page 135 InSQLBooks.chm: "Importing Data from an InTouch History File" topic |
-- |
Add the following bullet item: "If you have a large amount of data to import, process the data in batches, starting from the most recent data to the oldest data. For example, you want to import one year's worth of InTouch history. Divide up the data so that one month of data is included in a batch. When you import the most recent batch, the utility automatically starts with the most recent block and then proceeds backwards. Then, import the second-most recent batch, and so on." |
| 5 |
InSQLConcepts.pdf: "IDAS Late Data Handling" section, page 60 InSQLBooks.chm: "IDAS Late Data Handling" topic |
-- |
Add the following bullet item under the data throughput rules: "Late data must not encompass more than six history blocks at any given point of time." |
The Wonderware Historian documentation set includes the following guides:
This guide provides information on installing the Wonderware Historian, including hardware and software requirements and migration instructions.
A PDF file for each of these guides is available on the Wonderware Historian installation CD. You can easily print information from the PDF files. The Wonderware Historian documentation is also provided as an online Help file, which can be accessed from the System Management Console management tool. An installation Help file is also available from within the installation program.
Note: Any information
contained in the FactorySuite System Administrator's Guide regarding
Wonderware Historian is superseded by the documents referenced above.
© 2008 Invensys Systems, Inc. All Rights Reserved.