Software Pursuits Support Center

SureSync Documentation

SureSync 6

SureSync 6 Version History

Important: SureSync 6 support ended March 26, 2014

SureSync 6.0.39 Known Issues

Version 6.0.39 was posted on November 26, 2013.

Please note: The SPIAgent will test 6.0.37 in this release. That is expected behavior, no changes have been made to the component.

SureSync 7 has been released. Please visit the SureSync 7.0 New Features page for details.

Return to top of page

SureSync 6.0.38 Known Issues

Version 6.0.38 was posted on May 23, 2013.

File Synchronization

An "index was outside the bounds of the array" error could occurring during an execution of a Schedule. This has been fixed in 6.0.39.
In rare instances when a change event occured at the same instance a folder scan started for a subfolder, the change event could be skipped because the job thought it would be processed by the scan. This has been fixed in 6.0.39.

Please note: The SPIAgent will test 6.0.37 in this release. That is expected behavior, no changes have been made to the component.

SureSync 7 has been released. Please visit the SureSync 7.0 New Features page for details.

Return to top of page

SureSync 6.0.37 Known Issues

Version 6.0.37 was posted on March 21, 2013.

File Synchronization

The warning: “System.Data.SqlClient.SqlException: Violation of PRIMARY KEY constraint ‘PK_FileActions_Rel_NodeId_FileName’. Cannot insert duplicate key in object ‘dbo.FileActions’. The statement has been terminated.” could occur in a multi-mirror. This has been fixed in 6.0.38.
When enountering slow SQL database performance, SureSync could encounter a SQL timeout that would cancel the Relation, Schedule or Monitor in question. The timeout value has been increase to 2 minutes. This has been fixed in 6.0.38.
Changes were made to address missed events in a change journal enabled Schedule. This has been fixed in 6.0.38.
In situations where copy security is used and paths inherit different sets of security it was possible for rename events to be processed as a delete/add instead of a rename. This has been fixed in 6.0.38.

Real-Time Monitors

When running a new monitor with a new Relation, if a path was lost after the completion of the first full scan but before the monitor had been stopped/started a full rescan could incorrectly occur on path recovery. This has been fixed in 6.0.38.

SyncLockStatus

In rare circumstances, the Communication Service in SyncLockStatus could encounter a deadlock that would hang the service. This has been fixed in 6.0.38.
The Licenses tab of the Communication Service Configuration Utility was incorrectly supressing a scroll bar making it difficult to activate/deactivate licenses in environments with a lot of licenses. This has been fixed in 6.0.38.

Return to top of page

SureSync 6.0.36 Known Issues

Version 6.0.36 was posted on March 6, 2013.

File Locking

With the File Locking Add-on, changes were made to prevent situations where locks were either incorrectly held onto when they shouldn’t have been or were released prematurely. This has been fixed in 6.0.37.

File Synchronization

“Scan was interrupted. SelfGenChangeEventsCount” messages could occur that haulting processing of files. This has been fixed in 6.0.37.
In rare timing related situations, a manually run Relation or Schedule could handle renames inaccurately due to queuing of rename events into the queue table at the end of the job execution. This has been fixed in 6.0.37.

SPIAgent

It was possible to receive large numbers of “SPIAgentAsyncSend unable to locate target SPIAgent” messages in the Event Viewer that would eventually cause the SPIAgent service to stop to avoid an error loop. This has been fixed in 6.0.37.

SyncLockStatus

A refresh of shared components between SureSync and SyncLockStatus. It is recommended to upgrade the SyncLockStatus client in your environment when possible.

Return to top of page

SureSync 6.0.35 Known Issues

Version 6.0.35 was posted on November 13, 2012.

IMPORTANT: Please note that 6.0.36 includes a new version of the SPIAgent File Locking Add-on file filter driver. This will require a reboot of each machine when you’re applying the upgrade if your machines are running with the file locking Add-on. All machines using the file locking Add-on must be upgraded at the same time before restarting the locking Monitor.

File Locking

With the SPIAgent File Locking Add-on, in rare circumstances locks could be duplicated and then both locks not released. This would result in a file remaining locked when it was no longer open by a user. This has been fixed in 6.0.36.
With the SPIAgent File Locking Add-on, it was possible for the Detailed Status display to show a different number in the “Files Locked” counter compared to the actual list of file locks. This has been fixed in 6.0.36.
With the SPIAgent File Locking Add-on, corrections have been made to address situations where locks were not always released correctly. This has been fixed in 6.0.36.
Changes were made to the file locking driver to help prevent bluescreens. This has been fixed in 6.0.36.

File Synchronization

“Object reference not set to an instance of an object” error messages could occur when running a Relation, Schedule or Real-Time Monitor. This has been fixed in 6.0.36.
In some instances, the folder truncate delete function was not being used when it would be effective. This has been fixed in 6.0.36.
The error “an operating system subroutine returned an error. SPIVolMon::GetIoResultGetOverlappedResult error=996 (0x3e4)” error could occur when running a synchornization. This has been fixed in 6.0.36.
The code for handling renames has been optimized to address situations where a copy and paste action would result in an add/delete behavior instead of a rename. This has been fixed in 6.0.36.
A performance enhancement was made to increase the speed of a full folder scan when the folder structure was very deep and contained huge numbers of folders. This has been fixed in 6.0.36.
In rare circumstances, when a path was dropped from a Monitor and then quickly recovered cached events could be incorrectly processed resulting in a deleted file if the path in question was a source path. This has been fixed in 6.0.36.

Real-Time Monitors

Real-Time Monitors could be stopped by a “JrnWorkQueueIndexes entry not found for JrnlItemNumber=**” error message. This has been fixed in 6.0.36.

Schedules

It was possible for a schedule to occasionally hang on “waiting for file changes.” This has been fixed in 6.0.36.

SyncLockStatus

A correction was made to the SyncLockStatus installer to correctly write a registry key required when deploying the client with the /XMLPath switch. Previously, the registry key value was getting wiped out and recorded blank which broke the ability to deploy with /XMLPath. This has been fixed in 6.0.36.

User Interface

A change has been made to allow for a user to create a Relation with an SPIAgent path that is inaccessible. Previously, the save would not be allowed. This allows a user to configure a Relation when a particular SPIAgent is offline. This has been fixed in 6.0.36.
When trying to launch the local SPIAgent Configuration Utility from the Tools menu, the user would receive an “Unexpected error when launching a program. File not found.” error message. This has been fixed in 6.0.36.
A change was made to the Remote Installation Utility to allow backslash characters in the switches textbox. This is required for some SyncLockStatus distributions. This has been fixed in 6.0.36.

Return to top of page

SureSync 6.0.34 Known Issues

Version 6.0.34 was posted on August 13, 2012.

IMPORTANT: Please note that 6.0.35 includes a new version of the SPIAgent File Locking Add-on file filter driver. This will require a reboot of each machine when you’re applying the upgrade if your machines are running with the file locking Add-on. All machines using the file locking Add-on must be upgraded at the same time before restarting the locking Monitor.

File Locking

When using File Locking with PowerPoint ppt and pptx files, the PowerPoint files were not always retaining their locks when open. This has been fixed in 6.0.35.
When using File Locking, a delete of a folder on one side of the multi-mirror would result in the contents of the folder being deleted on the other side(s) but not the folder itself. This left an empty folder behind that should have been deleted. This has been fixed in 6.0.35.
When using File Locking, if a folder delete was processed while there were outstanding locks for files in the folder it was possible for an empty folder to be left behind. This has been fixed in 6.0.35.
When using File Locking, if a user had files open before the monitor was started and then closed the file a lock would be generated on the close and not released. This has been fixed in 6.0.35.
The testing of lock closes has been moved to a separate thread which enhances file locking performance when under heavy load. This has been fixed in 6.0.35.
Performance optimizations have been made to the SPIAgent File Locking driver. This has been fixed in 6.0.35.
Changes were made to address an obscure bluescreen that was possible in rare situations with the file locking add-on driver. This has been fixed in 6.0.35.

File Synchronization

If a job was terminated while active file actions were queued and other events from later in the journal had already been processed, the journal location could be improperly check pointed. This could cause events for those active file copies to be skipped when the job was restarted. This would generally only affect schedules. This has been fixed in 6.0.35.
When using a mapped drive in a Schedule, it was possible for the Schedule to not always execute the full scan. This could result in files not being synchronized until the next full scan occurred. This has been fixed in 6.0.35.
A “Cannot set column ‘Subfolder’. The value violates the MaxLength limit of this column.” error could occur when running a job with the “write file statistics in JobStatistics table” option enabled. This has been fixed in 6.0.35.
When processing a large number of folder deletes occurring at the same time, it was possible for the contents of the folders to be deleted but not the folders themselves. This has been fixed in 6.0.35.
Windows 8 and Windows Server 2012 are now supported with the exception of File Locking. File Locking support for these operating systems will be included in SureSync 7. This has been fixed in 6.0.35.
Temporary ~~SureSync.tmp folders could be left behind if the folders contained obsolete files that no longer exist in the file structure. A full scan is required for the cleanup of these folders to occur. This has been fixed in 6.0.35.
If you alter security values on a folder or file or create a new folder with non-inherited permissions and your have multiple destinations, the security values may not be copied to all destinations. This has been fixed in 6.0.35.

Scheduler

If a server or agent reference was deleted using the Desktop it was possible for the active scheduler using the database to go into an error loop and stop selecting jobs. The error wouldn’t persist if the Scheduler was restarted. This has been fixed in 6.0.35.

SyncLockStatus

On machines running the .NET 4.5 Framework, the SyncLockStatus client could consume 20% or more CPU usage due to a bug that Microsoft has identified in the .NET 4.5 Framework. One of the workarounds provided by Microsoft has been implemented to attempt to avoid the problem. This has been fixed in 6.0.35.

User Interface

When running a Preview, it was possible to encounter an “object reference not set to an instance of an object” error message. This has been fixed in 6.0.35.
If you create a one way Rule with three or more paths and then select only one path as source and one path as destination, the desktop would incorrectly prompt to delete the Rule when you deleted one of the paths not selected as source or destination. This has been fixed in 6.0.35.
If a Relation was configured with two root paths, the user deleted the destination path, added a new destination path, and then modified the source path to a new path (all at once without clicking the apply button) then the desktop could produce errors and not display the new paths. This has been fixed in 6.0.35.
When loading an extremely large job log in the Jog Log Viewer, the memory usage of SQL could bloat due to an error in one of the job log viewer queries. This has been fixed in 6.0.35.

Return to top of page

SureSync 6.0.33 Known Issues

Version 6.0.33 was posted on August 1, 2012.

File Locking

File locking was incorrectly suppressed when using an Exchanging Copies rule. This has been fixed in 6.0.34.

File Synchronization

If you configured a root path that contains underscores in a UNC path job, it was possible for the path to be incorrectly recognized as a Netware path. This would cause the path to be passed incorrectly and lead to wrongfully identifying the server associated with the UNC path. This could cause a licensing error. This has been fixed in 6.0.34.
When executing a Schedule that is configured with UNC paths, it was possible for the Schedule to not always execute a full scan. This could result in files not being synchronized until the next full scan occurred. This has been fixed in 6.0.34.

Scheduler

When using multiple Schedulers or executing Schedules with a single Scheduler that complete in less than 5 minutes, it was possible to receive alerts indicating that the Schedule had missed its scheduled execution date and time when it had not. This has been fixed in 6.0.34.

SPIAgent

If the primary network connection on a machine is a wireless connection, it was possible for the SPIAgent service to initialize before the machine had an IP address due to the slowness of IP address assignment on some wireless interfaces. This would result in the SPIAgent on that machine not working. This has been fixed in 6.0.34.
A change has been made to the SPIAgent to allow it to try each port in the SPIAgent port range when certain TCP errors are encountered obtaining a port. Previously the SPIAgent would abort on these errors. This has been fixed in 6.0.34.

Return to top of page

SureSync 6.0.32 Known Issues

Version 6.0.32 was posted on July 13, 2012.

IMPORTANT: Please note that 6.0.33 includes a new version of the SPIAgent File Locking Add-on file filter driver. This will require a reboot of each machine when you’re applying the upgrade if your machines are running with the file locking Add-on. All machines using the file locking Add-on must be upgraded at the same time before restarting the locking Monitor.

Database

In environments running versions of SureSync older than 6.0.31, if a SPIAgent machine was upgraded before the machines running the full SureSync software it was possible for the database to become corrupted. The corruption will be repaired automatically without any harm to the database by upgrading the SureSync machine to 6.0.31 or newer. A change has been made to restore the backwards compatibility to avoid this issue. This has been fixed in 6.0.33.

Desktop

In an environment where SureSync is licensed for a server, a server SPIAgent, and workstation SPIAgent licenses the license panel was incorrectly displaying Temp License Days for the workstation agents when activations are not required for those agents. This is a display bug and does not affect operation of the program. This has been fixed in 6.0.33.
In SPIAgent Setup, if a SPIAgent name was entered into the drop-down list that included the \ character then a “Unable to cast object of type ‘System.Int32’ to type ‘SPIServerSetLib.SPIServerItem’. An internal application error has occurred” message would be produced and the ability to edit machines in SPIAgent Setup would become disabled. This has been fixed in 6.0.33.

File Locking

The File Locking Add-on was incorrectly allowing a user to delete a read-only copy of a file that should have been rejected because the file was locked on another machine involved in the Real-Time Monitor. This has been fixed in 6.033.

File Synchronization

Enhancements have been made to the SureSync Scheduler service to enhance scheduling accuracy in multiple Scheduler environments. When using multiple Schedulers, the rescheduling interval for Schedules must be 5 minutes or longer for accurate rescheduling due to overhead associated with multiple Schedulers reading and updating the SureSync database. This has been fixed in 6.0.33.
When renaming a file, the error “System.Data.ConstraintException: Column ‘RelationId,FileId,RootNumber’ is constrained to be unique” could occur and the error would cancel the job. This has been fixed in 6.0.33.
When renaming a folder, the error “This row has been removed from a table and does not have any data” could occur and the error would cancel the job. This has been fixed in 6.0.33.
When renaming folders or files, the error “Violation of primary key constraint” could occur and the error would cancel the job. This has been fixed in 6.0.33.
When using a Schedule or Monitor with Change Journal support, the error “Invalid input parameter received: SVR.StartUSN must be an even number” could be encountered. This happens because SureSync received a bad USN number from the Change Journal. The job will now reset the bad value and force a rescan of the directory structure to resolve the issue. This has been fixed in 6.0.33.

Real-Time Monitors

When running a Relation with many paths offline at once in a Real-Time Monitor, the monitor would appear to be hung when it was really processing retry logic. Changes have been made to allow other events through and avoid the appearance of a hang. This has been fixed in 6.0.33.

SyncLockStatus

A refresh of shared components between SureSync and SyncLockStatus. It is recommended to upgrade the SyncLockStatus client in your environment when possible.

Return to top of page

SureSync 6.0.31 Known Issues

Version 6.0.31 was posted on May 23, 2012.

File Locking

When running a Real-Time Monitor with File Locking enabled, it was possible for a retry event to be queued in the database that lead to a “collection was modified after the enumerator was instantiated” error that would cancel the Real-Time Monitor. This has been fixed in 6.0.32.
When running a Real-Time Monitor with File Locking enabled, if a user moved a large number of files from one monitored folder to another, it was possible for some of the files to be duplicated back into the original directory while the files were being processed. This has been fixed in 6.0.32.

File Synchronization

When launching the SureSync desktop user interface it was possible to receive an “Index was outside the bounds of the array. An internal application error has occurred.” error. This error could also be reported on the launch of a Relation, Schedule or Real-Time Monitor. This has been fixed in 6.0.32.
With a Schedule running a move Rule, it was possible to encounter a situation where the source file could not be deleted after the successful move. This event was incorrectly classified and not retried which resulted in the source file not being deleted. This has been fixed in 6.0.32.
When running multiple Schedulers, it was possible to receive messages from a Scheduler saying that a Schedule missed its scheduled execution time. This message was incorrect and the Scheduler was actually starting the Schedule. This has been fixed in 6.0.32.
On some non-Windows emulated NTFS file systems, it appears that the file system modifies the time and date stamp when renaming a file from SureSync’s temporary folder resulting in unnecessary copying. A change has been made to set the timestamp after the rename instead of before to avoid the problem. This has been fixed in 6.0.32.
On some non-Windows emulated NTFS file systems, file names can be forced to all lower case characters. If the source files had upper case characters the SureSync job would try to rename the files to the correct case with each run. An option has been added to Path Options to ignore case differences. This has been fixed in 6.0.32.
It was possible to have a license marked in use when it was not resulting in messages about exceeding licenses. Generally, this involved adding a reference to an incorrect machine and then changing it. This has been fixed in 6.0.32.
On some non-Windows paths, it was possible for the Path Access user ID defined in Path Options to have the credentials timeout with that non-Windows operating system. This would result in a “logon failure: unknown username or bad password” error. SureSync now refreshes the credentials more frequently to avoid the problem. This has been fixed in 6.0.32.
Security changes made with “copy folder security objects to existing folders” and “copy file security objects even if the file is not copied” could result in a timeout when a security change was made that impacted a large section of a big folder structure. A “SyncFiles job terminated immediately due to unresponsive local agent” message in the Application Event Log would be generated. The timeout has been doubled to an hour. Additional increases can be specified in the database if necessary, contact technical support for details. This has been fixed in 6.0.32.
When running a Schedule or Real-Time Monitor it was possible to receive a “Timeout expired. The timeout period elapsed prior to the completion of the operation or the server is not responding” message that canceled the job. This was possible with a slow SQL server and the timeout has been increased to address the issue. This has been fixed in 6.0.32.
The handling of folder deletes for multi-way Rules has been modified to not use a truncate delete. This type of delete will delete the entire folder on the destination without processing the contents. Instead, it will now delete each subfolder and file separately before deleting the parent. If excluding hidden and system files, errors about folders not being empty will be possible because each file is checked by the Rule. A new option on the Options tab of a Relation can turn truncate deletes on or off for one-way Rules. This has been fixed in 6.0.32.

Preview

When launching a Preview and clicking on the Filter tab, all of the paths were incorrectly listed with the named agent of the machine launching the Preview instead of the correct agent name for each path. This has been fixed in 6.0.32.

SPIAgent

In some circumstances, it was possible for the SPIAgent service to terminate unexpectedly. This occurred when the SPIAgent encountered 500 or more Access Denied messages on folders. These events would be logged in the Application Event Viewer. This has been fixed in 6.0.32.

SyncLockStatus

A refresh of shared components between SureSync and SyncLockStatus. It is recommended to upgrade the SyncLockStatus client in your environment when possible.

Return to top of page

SureSync 6.0.30 Known Issues

Version 6.0.30 was posted on April 30, 2012.

File Synchronization

Advanced Open File Support via VSS was not functioning properly due to an error in the Software Pursuits VSS .dll files. This has been fixed in 6.0.31.

File Locking

With SPIAgent File Locking, it was possible for a file to remain locked after close due to the file being stuck in a retry sequence. This has been fixed in 6.0.31.
On Windows 2008 64-bit machines, it was possible for the file locking driver to fail to load due to a signing issue on the SPIFMon.sys component. This has been fixed in 6.0.31.

SPIAgent

A 900,000 millisecond timeout error could occur on AReqType_ShowStats that could result in a dropped path. This has been fixed in 6.0.31.

SyncLockStatus

A refresh of shared components between SureSync and SyncLockStatus. It is recommended to upgrade the SyncLockStatus client in your environment when possible.

Return to top of page

SureSync 6.0.29 Known Issues

Version 6.0.29 was posted on April 26, 2012.

SPIAgent

It was possible to receive a “could not load file or assembly ‘SPIAgent6.dll’ or one of its dependencies. This assembly is built by a runtime newer than the current loaded runtime and cannot be loaded.” when loading the SureSync desktop or starting the SPIAgent service. This has been fixed in 6.0.30.

SyncLockStatus

A refresh of shared components between SureSync and SyncLockStatus. It is recommended to upgrade the SyncLockStatus client in your environment when possible.

Return to top of page

SureSync 6.0.28 Known Issues

Version 6.0.28 was posted on March 1, 2012.

Database

A “QItemSubpath exceed maximum size” error could occur when copying a SQL SureSync database when that database contains a retry path in the database that is longer than 255 characters. This has been fixed in 6.0.29.
Changes have been made to the SureSync database upgrade/copy routine to allow the user to copy into an existing empty database. Database professionals can now create a blank database on a SQL server and grand db_owner privileges to the SureSync account. The SureSync administrator can then copy into the existing database. This is only an issue in high security environments. This has been fixed in 6.0.29.

File Synchronization

Changes have been made to the FTP support to allow the use of an absolute path instead of relative paths. This option is defined on a per path basis by editing the path, clicking on the Options tab in the “Edit Root Path,” dialog box and checking the “FTP path is absolute” option. This has been fixed in 6.0.29.
A Relation, Schedule or Monitor could hang on shutdown when encountering 500 of the same error in a row. When this condition occurs, the job is canceled to prevent an error loop but was not always happening. This has been fixed in 6.0.29.
If a UNC path specified a server name of longer than 30 characters, the server name would not be properly recognized by the licensing system. This has been fixed in 6.0.29.

Real-Time Monitors

An “object reference not set to an instance of an object” error could cancel a Real-Time Monitor. This error would occur if some history was missing for a path. This has been fixed in 6.0.29.
If a Real-Time Monitor with File Locking enabled lost a connection in the middle of an active file transfer, the file transfer would get queued for retry but not processed. This has been fixed in 6.0.29.

SyncLockStatus

A refresh of shared components between SureSync and SyncLockStatus. It is recommended to upgrade the SyncLockStatus client in your environment when possible.

User Interface

It was possible to receive a SureSync Path Info error 122 ()x7a) SPIAgent Return Code-APathInfoError in browse. This was due to a buffer being used to convert drive letters to UNC paths being too small. This has been fixed in 6.0.29.

Return to top of page

SureSync 6.0.27 Known Issues

Version 6.0.27 was posted on February 2, 2012.

Desktop

If you put a Relation, Schedule, or Real-Time Monitor in a non-default group folder and assign special permissions to the folder while defining to block inherited permissions then you would lose access to that item. This has been fixed in 6.0.28.
The SureSync desktop was allowing some server references in Tools | SPIAgent Setup to be removed when those servers were in use. This would lead to a license count problem. This has been fixed in 6.0.28.

File Synchronization

Incorrect syntax near ‘,’. Must declare the scalar variable “@RelationID” errors could occur due to a syntax error in a SQL statement that is executed when a Relation completes and items are in the retry queue in the database. This has been fixed in 6.0.28.
“Error setting file attributes. The system cannot find the file specified.” errors could occur when setting attributes on a file. This has been fixed in 6.0.28.
If you set the execution priority to the lowest setting, that setting was not used and the job was executing at a normal execution priority. This has been fixed in 6.0.28.

Scheduler

If you try to start a number of Schedules at the same time and you are running multiple Schedulers, some Schedules may not start on time. Since some Schedules could be delayed, they could miss their scheduled run time and generate an error message. This has been fixed in 6.0.28.

Return to top of page

SureSync 6.0.26 Known Issues

Version 6.0.26 was posted on December 2, 2011.

ArchiveSync

If you have obsolete archive entries and used the option to delete items when no rule applies, the files may be reported as deleted from the archive when in fact they were not. This has been fixed in 6.0.27.

Databases

Errors were found in copying SureSync databases to SQL Server and in purging matching retry entries in SQL Server. Paths longer than 400 characters may not have been retried properly. you might also get an error message reporting that the data was too long for a column in the Queue table. This has been fixed in 6.0.27.

File Synchronization

When using the “Copy file when source file is open” option and not “Use Advanced Open File Support”, a retry for an open file copy might only be processed once when a Schedule is used. A re-scan could find an unprocessed file. If you are running a Schedule configured in this manner, a full rescan is recommended after upgrading. This has been fixed in 6.0.27.

SyncLockStatus

A refresh of shared components between SureSync and SyncLockStatus. This has been fixed in 6.0.27.
In some circumstances, the SyncLockStatus client would incorrectly present messages indicating that a file was available when it wasn’t. This has been fixed in 6.0.27.

Return to top of page

SureSync 6.0.25 Known Issues

Version 6.0.25 was posted on October 14, 2011.

IMPORTANT: Please note that 6.0.26 includes a new version of the SPIAgent File Locking Add-on file filter driver. This will require a reboot of each machine when you’re applying the upgrade if your machines are running with the file locking Add-on. All machines using the file locking Add-on must be upgraded at the same time before restarting the locking Monitor.

Desktop Interface

On some foreign language editions of Windows, the error “Invalid property array indexAn internal application error has occurred” could occur when working through the Schedule wizard. This has been fixed in 6.0.26.
Changes have been made to the SureSync desktop user interface to address Relations, Schedules, and Monitors that would appear stuck in an “active” state when not actually active. This has been fixed in 6.0.26.

File Archiving

In an archive job, if you archived files or folders and changed the case of characters in the names of those items, the case change was not properly recorded in the archive. This lead to an attempt to re-archive the file or folder each time an event is handled for the file or each time a Schedule is run. This has been fixed in 6.0.26.
With an archive job, archive thinning is queued after file actions complete, but Schedules could end before the thinning operation completed. This could leave more items in the archive than indicated by the archive rule. A full rescan (right click on the Schedule or Monitor and select “Re-scan files”) is recommended on any ArchiveSync schedules to force an archive thinning. This has been fixed in 6.0.26.

File Locking

In rare circumstances, a Monitor running with File Locking could encounter a timeout that could cause slow response time from the server or the impression of a hang. Stopping the job or waiting for the timeout to occur would resolve the issue. This has been fixed in 6.0.26.

File Synchronization

When running a synchronization with a FTP path, the message “Rebex.Net.FtpException: Timeout exceeded” could occur in invalid instances with some FTP servers. This has been fixed in 6.0.26.
When running a job that uses change journal support and a system had an extremely large volume, the message “Invalid input parameter received: SVR.StartUSN must be even number” could occur that would cancel the job. This has been fixed in 6.0.26.
When creating nested folders from a drive letter or file share, an incorrect test was made against the drive letter or share to see if it existed. This would result in an “error trying to find folder” message that prevented creating the needed folder. This has been fixed in 6.0.26.
When the read-only attribute was set on a file on one side of a synchronization before a rename occurred on the same file, that file was not processed. The file was only processed on a full rescan. This has been fixed in 6.0.26.
If a path was down when running a Schedule, in some instances that Schedule would hang on “Draining Change Journal”. This has been fixed in 6.0.26.

Real-Time Monitor

When running a monitor that involved a path with an extremely large volume, the error “an unexpected program exception has occurred” could be produced by SyncFiles.exe due to a buffer in the SPIAgent that was too small. This would cause the monitor to be placed on hold. This has been fixed in 6.0.26.

SyncLockStatus

SyncLockStatus was refreshed to use the current shared components. If you are running previous versions of SyncLockStatus, you should not encounter any problems receiving status from SyncLockStatus on the workstation machines. An upgrade is recommended when convenient. This has been fixed in 6.0.26.

Return to top of page

SureSync 6.0.24 Known Issues

Version 6.0.24 was posted on August 30, 2011.

IMPORTANT: If your environment is running file locking, you must upgrade all machines running the SPIAgent to the 6.0.25 release level at the same time. Changes have been made to file locking that are incompatible with previous releases.

File Locking

When a large number of locks are present, performance could be degraded in processing files because of internal tests to see if locks should be released. The duration between these tests has been extended and unnecessary tests have been eliminated. This has been fixed in 6.0.25.
When file locking was enabled in a real-time monitor, in some rare timing instances, a lock release could be triggered before the file copy. This out of order request kept the file delayed in the retry queue and the file was not copied. A restart of the monitor would trigger the retries queued to be processed. This has been fixed in 6.0.25.

Real-Time Monitors

A file rename in a multi-mirror could, on rare occassions, cancel the job if partial file history was present at the time of the rename. The error canceling the job would be Violation of PRIMARY KEY constraint ‘PK_FileHistory_Rel_FileId_RootNumber’. Cannot insert duplicate key in object ‘dbo.FileHistory’. This has been fixed in 6.0.25.
When individual change journal events were processed with a multi-way rule, processing could be degraded, adding about 100 milliseconds to each file processes. This has been fixed in 6.0.25.

SyncLockStatus

SyncLockStatus was refreshed to use the current shared components. If you are running previous versions of SyncLockStatus, you should not encounter any problems receiving status from SyncLockStatus on the workstation machines. An upgrade is recommended when convenient. This has been fixed in 6.0.25.

Return to top of page

SureSync 6.0.23 Known Issues

Version 6.0.23 was posted on August 16, 2011.

File Synchronization

UNC path jobs would incorrectly drop a UNC path that was online. If the job contained only two paths, this would make the entire job fail due to lack of sufficient root paths. This has been fixed in 6.0.24.

SyncLockStatus

SyncLockStatus was refreshed to use the current shared components. If you are running previous versions of SyncLockStatus, you should not encounter any problems receiving status from SyncLockStatus on the workstation machines. An upgrade is recommended when convenient. This has been fixed in 6.0.24.

Return to top of page

SureSync 6.0.22 Known Issues

Version 6.0.22 was posted on June 7, 2011.

Database

In rare circumstances, a version 5.0 database could contain orphaned records that conflicted with the database integrity rules. This would result in a “the inserted statement conflicted with the foreign key constraint” error message during a database upgrade. Changes were made to the conversion to clean up the orphaned records. This has been fixed in 6.0.23.

Desktop

In the SureSync desktop user interface, if the mouse was hovering over the tree of jobs when a refresh occurred it was possible to encounter a “System.NullReferenceException: Object reference not set to an instance of an object” error. This has been fixed in 6.0.23.

File Locking

Memory mapped files, such as some PowerPoint files, could have file locking locks released prematurely. Microsoft Office files that create temporary files prefixed with “~$” will now be locked as a pair. This has been fixed in 6.0.23.

File Synchronization

If a job contains a multi-mirror rule, multiple destination paths, a path is offline, and a rename is performed then it was possible to receive a PRIMARY KEY constraint ‘PK_FileHistory_Rel_FileId_RootNumber’ error. This has been fixed in 6.0.23.
When a job running with Change Journal support had many pending Change Journal events to be processed, it was possible for the job to get a timeout when trying to detect if a connection is available. The timeout would cause the path to drop. This has been fixed in 6.0.23.
Filtering on attributes has been enhanced to consider folder attributes. On the Attributes tab of a rule, selecting “Use attributes for selecting folders” will tell the Rule to consider folder attributes. This is useful for filtering out system and hidden folders. Existing jobs will not be modified but new jobs will consider folder attributes by default. This has been changed in 6.0.23.

Real-Time Monitors

When filtering by file size in a Real-Time Monitor, files that are to be delayed could be retried every 15 seconds instead of being delayed until the next interval. This has been fixed in 6.0.23.
When configuring a Real-Time Monitor with a mirror rule to monitor an NTFS source directory (with Change Journal support) to a UNC path (no Change Journal) and the option in Path Options to monitor the UNC past was not unchecked then the job was failing to warn the user about having to turn that option off. This resulted in files not being synchronized. This has been fixed in 6.0.23.

SyncLockStatus

SyncLockStatus was refreshed to use the current shared components. If you are running previous versions of SyncLockStatus, you should not encounter any problems receiving status from SyncLockStatus on the workstation machines. An upgrade is recommended when convenient. This has been fixed in 6.0.23.

Return to top of page

SureSync 6.0.21 Known Issues

Version 6.0.21 was posted on April 11, 2011.

File Synchronization

Duplicate entries could appear in the “Retries Queued” counter. This has been fixed in 6.0.22.
In some circumstances, change journal events could be read multiple times. This was harmless but a set of changes have been made to further optimize reading the change journal. This has been fixed in 6.0.22.
E-mail alerts set to authenticate anonymously with the defined mail server would sometimes fail to send depending on e-mail server settings. This would result in the alert not being sent. This has been fixed in 6.0.22.

SPIAgent

When using the SPIAgent, “unable to read from source file” and “invalid file pointer” messages could occuring while processing a file. This has been fixed in 6.0.22.

SPIAgent File Locking

In rare instances, a send request sending a lock message to a remote agent could take an excessive amount of time to complete. This would cause the locking session to be stuck waiting for a response and other lock messages could time out. This could make a file appear locked when it was not. This has been fixed in 6.0.22.

SyncLockStatus

If a workstation existed in a workgroup, it would fail to connect to a Communications Service located in a domain. This has been fixed in 6.0.22.

User Interface

When refreshing a job log entry in the job log viewer, the currently selected log would not be selected again after the refresh. This required the user to select the log from the drop-down list again. This has been fixed in 6.0.22.

Return to top of page

SureSync 6.0.20 Known Issues

Version 6.0.20 was posted on February 18, 2011.

File Synchronization

When using a Move Rule, if a source folder is created and the same folder exists on the destinations, it was possible for the source folder to be deleted even when it contained other files. In this situation, the files in the source folder were created after the folder was created and before the attempted move of the folder. The folder create event triggered the source folder delete since it already existed on the destinations. This has been fixed in 6.0.21.
In rare circumstances where a preview had been performed and then a Rule was deleted from the Relation, the remaining file history could cause a “System.IndexOutOfRangeException: Index was outside the bounds of the array” error that would cancel the job. This has been fixed in 6.0.21.
In certain circumstances, an internal error could occur attempting to cancel a file copy when the user cancels a job. This would cause the job to fail and not respond to other commands or requests for status. This has been fixed in 6.0.21.
Relations using a SFTP path could incorrectly fail with an “Unable to access Root Path. System.IO.DirectoryNotFoundException error. This has been fixed in 6.0.21.
If a job is configured to run with a single thread and that thread aborts because a connection to an SPIAgent is lost while copying files and you have 50 requests queued, your job may hang. This has been fixed in 6.0.21.
In rare timing conditions when deltas are used, an internal error could occur when re-positioning a file being read. An erroneous error may be generated, canceling the file copy. This has been fixed in 6.0.21.

User Interface

Arithmetic operation resulted in an overflow errors could occur in Detailed Status when processing more than 2 billion journal events. This has been fixed in 6.0.21.

Real-Time Monitors

When using a relative date range with a Real-Time Monitor, files could be replicated immediately upon modification and ignore the date tab setting. This has been fixed in 6.0.21.
When using Advanced Open File Support in a Real-Time Monitor, comments on a file’s properties could not be synchronized until another journal event came through processing a re-copy of the file. This has been fixed in 6.0.21.
If using Real-Time Monitors with gaps in the Intervals, the gap could be ignored and the “Between Intervals” status missed. This would result in files being synchronized when processing should have been paused. This has been fixed in 6.0.21.

SyncLockStatus

SyncLockStatus was refreshed to use the current shared components. If you are running previous versions of SyncLockStatus, you should not encounter any problems receiving status from SyncLockStatus on the workstation machines. An upgrade is recommended when convenient. This has been fixed in 6.0.21.

Return to top of page

SureSync 6.0.19 Known Issues

Version 6.0.19 was posted on January 3, 2011.

File Synchronization

A NullReferenceException error could be generated if an SPIAgent was not available while a job was starting up and was later recovered into the job. This has been fixed in 6.0.20.
Further enhancements have been made to the bandwidth throttling functions on a Priority to enhance throttle accuracy. This has been fixed in 6.0.20.
In certain rare circumstances, a “this row has been removed from a table and does not have any data” or other database error messages may be generated for a multi-mirror job following a folder rename in some cases. This has been fixed in 6.0.20.
If a source file changes while a file copy is in progress, those events could be ignored. This could prevent the final status change for a file to be recognized. This has been fixed in 6.0.20.
An “Arithmetic Overflow” error could occur if a job processes over 2 billion Change Journal events or, possibly, if a connection is lost and recovered. This would generate multiple errors before canceling the job. This has been fixed in 6.0.20.

Real-Time Monitors

In certain rare circumstances, a record in the Change Journal could be skipped. If that record is for a monitored file, actions for that file could be skipped until a full scan occurred. This has been fixed in 6.0.20.
If a file copy was being processed to multiple subnets, a loop could occur that would cause a Real-Time Monitor to hang and stop processing files. This has been fixed in 6.0.20.

SyncLockStatus

SyncLockStatus was refreshed to use the current shared components. If you are running 6.0.19 of SyncLockStatus, you should not encounter any problems receiving status from SyncLockStatus on the workstation machines. An upgrade is recommended when convenient. This has been fixed in 6.0.20.

Return to top of page

SureSync 6.0.18 Known Issues

Version 6.0.18 was posted on November 23, 2010.

ArchiveSync

When running an archive Restore Relation, the Relation could encounter a loop when processing a folder that was not selected by the Rule. This most likely occurs with date exclusions. This has been fixed in 6.0.19.

File Synchronization

Enhancements have been made to the bandwidth throttling functions on a Priority to enhance throttle accuracy. This has been fixed in 6.0.19.

Real-Time Monitors

When running multiple Real-Time Monitors, a Real-Time Monitor could cancel and be placed on hold with the error “SPIException Error #1024: System.IO.IOException: The process cannot access the file”. This has been fixed in 6.0.19.
When running a Real-Time Monitor with a relative date range defined on the Dates tab of the Rule, files skipped would not be retried when the file finally fell within the dates that should be processed. This resulted in files being missed. This has been fixed in 6.0.19.
When running a Real-Time Monitor, a “SPIException Error #1059: System.Data.RowNotInTableException: This row has been removed from the table or does not have any data. BeginEdit() will allow creation of new data in this row” error could occur that would place the Real-Time Monitor on hold. This has been fixed in 6.0.19.

User Interface

The Detailed Status panel was not always showing the correct count for retries queued. The display will now show the retries that are queued for retry within the next 60 minutes. This has been fixed in 6.0.19.
When a job terminated incorrectly, the SureSync desktop could still show the job as busy. A refresh of the desktop would not clear the busy status. Some attempts to clear busy status on the SureSync desktop were failing. This has been fixed in 6.0.19.

Return to top of page

SureSync 6.0.17 Known Issues

Version 6.0.17 was posted on October 27, 2010.

File Locking

In certain timing related instances, a file lock on a remote SPIAgent machine could not be processed which allowed users on other machines access to the same file. This has been fixed in 6.0.18.

File Synchronization

In certain circumstances, files could be skipped due to an “unable to read from source file” error. This has been fixed in 6.0.18.
The Windows Event Log could show a job crash error logged on termination of a job. The error is harmless but could clutter the log. This has been fixed in 6.0.18.
In certain circumstances, a “destination file is open by another user” error could occur that actually referred to a locked source file. In this case, the reporting was incorrect. This has been fixed in 6.0.18.
A SyncFiles.exe process could encounter a loop and consume 100% of a CPU core when encountering an error with a file resume. This has been fixed in 6.0.18.

Real-Time Monitors

In a monitor with more than two paths, a SPIAgent service being restarted on one machine could cause multiple path drops to be reported. These paths would be recovered on the next retry. This has been fixed in 6.0.18.
If you start a Monitor that has more than two paths, with one those paths down, the error recovery would recover the path that was down and immediately drop it again. It would then recover again on the next retry and be fine. The second path drop was incorrect and invalid. This has been fixed in 6.0.18.
A Real-Time Monitor could encounter a “System.NullReferenceException: Object reference not set to an instance of an object” error and then go on hold. This has been fixed in 6.0.18.

SPIAgent

It was possible to have direct SPIAgent-to-SPIAgent communications occur between two remote SPIAgent machines when the option was not set to allow it to do so. This has been fixed in 6.0.18.

User Interface

When launching Enterprise Status or Detailed Status, the error “System.OverflowException: Arithmetic operation resulted in an overflow” could occur. This has been fixed in 6.0.18.

Return to top of page

SureSync 6.0.16 Known Issues

Version 6.0.16 was posted on October 1, 2010.

Enterprise Status

In the Enterprise Status panel, the scroll controls were not always being displayed correctly when the application was resized. This has been fixed in 6.0.17.

File Synchronization

When a Schedule was run using Change Journal support and had no events to process, the Schedule could unnecessarily read the same Change Journal entries again. This has been fixed in 6.0.17.

Job Log

Job log entries for Real-Time Monitors that run longer than the defined setting for purging old logs could be incorrectly purged. This has been fixed in 6.0.17.

Preview

SyncFiles would fail to terminate properly when closing a Preview panel. This has been fixed in 6.0.17.
Improvements have been made to the “maximum dynamic network load” bandwidth throttling option to make the throttling more accurate. This has been fixed in 6.0.17.
In some circumstances, a job could encounter a “Verify of write failed for destination” error message. This has been fixed in 6.0.17.

SPIAgent

The SPIAgent could report large amounts of “CloseSock close socket failure code 10038: WSAENOTSOCK” error messages from a source of SPIMsg in the Application Event Viewer. These messages were harmless but cluttered the log. This has been fixed in 6.0.17.
The error “an operating system subroutine returned an error. SPIVolMon::ReadMFT GetOverlappedResult error=38 (0x26) message could be recorded when reading a Change Journal. This has been fixed in 6.0.17.
If a server name does not match its “Agent Access Name”, which is normally a DNS name, a dropped path may be reported for other agents even though their connection was not lost. This has been fixed in 6.0.17.
900,000 millisecond timeout error messages could occur when using the SPIAgent. This would result in the job appearing to be unresponsive while the timeout was happening. This has been fixed in 6.0.17.

Return to top of page

SureSync 6.0.15 Known Issues

Version 6.0.15 was posted on August 23, 2010.

ArchiveSync

ArchiveSync archives were not always thinning folders properly until a rescan or other event occurred for the folder. This has been fixed in 6.0.16.

File Synchronization

A delay could occur when canceling a job, in some cases this delay could be as long as 30 seconds. This has been fixed in 6.0.16.
A job could be canceled when failing to send an e-mail alert. This has been fixed in 6.0.16.
Retry improvements have been made related to retries of a file with multiple destinations and retries related to encountering a busy destination file. This has been fixed in 6.0.16.
When using a delete rule with the folder attribute set and a date range or other restrictive selection, a folder could be deleted when it should be retained. This has been fixed in 6.0.16.
The Software Pursuits Scheduler Service had an unnecessary dependency on NetLogon service which prevented the Scheduler from starting on standalone machines or machines in a workgroup. This has been fixed in 6.0.16.
In some circumstances, temporary folders (~~SureSync.tmp folders) which should have been cleaned up during a folder scan were not. This has been fixed in 6.0.16.
In certain unusual timing conditions, a manual cancel of a job could cause the job to deadlock, preventing it from terminating. This has been fixed in 6.0.16.
A “Transaction (Process ID ###) was deadlocked on lock resources with another process and has been chosen as the deadlock victim. Rerun the transaction.” error could occur when performing actions such as purging the job log. Generally, this error is harmless. This has been fixed in 6.0.16.
If you quickly delete a folder and then quickly put it back in a multi-mirror, it is possible for the folder contents to be deleted from the restored folder. This has been fixed in 6.0.16.

Preview

Jobs would fail to launch with an “object reference not set to an instance of an object” error when launched from the Preview. This has been fixed in 6.0.16.

Real-Time Monitors

If a file is being processed due to a retry operation and during that retry a path is lost, the action for the file may not be tried again. This has been fixed in 6.0.16.
If a path is lost and restored within about 30 seconds the agent may retain a handle to a file that was actively being copied. When attempting to resume the file copy it will find the file is busy. That condition may remain until another file is successfully copied. This has been fixed in 6.0.16.
If a path is dropped during a folder scan that has not generated any actions, the scan may not be restarted when the path is restored. This has been fixed in 6.0.16.
A monitor with dropped paths may continue reading and processing change journal entries even when it is not possible for those entries to generate any productive work. This has been fixed in 6.0.16.
A Real-Time Monitor could be canceled and placed on hold due to an “object variable not set to an instance of an object” error when processing a folder rename. This has been fixed in 6.0.16.

SPIAgent

If both the source and destination in a file copy were remote from the controlling SureSync machine, delta copies could be incorrectly suppressed. This has been fixed in 6.0.16.
900,000 millisecond timeout error messages could occur when using the SPIAgent in rare circumstances when performing local requests. This has been fixed in 6.0.16.

SyncCmd

The SyncCmd /U command failed to update remote SPIAgent installation services as it should. This has been fixed in 6.0.16.

Return to top of page

SureSync 6.0.14 Known Issues

Version 6.0.14 was posted on July 6, 2010.

Desktop

The SureSync desktop refresh was not clearing all busy nodes. If you had a job which failed that contained multiple Relations, you could encounter a situation where a refresh of the Desktop only set one of the Relations to not active. Each additional refresh would reset the status of one more of the Relations. This has been fixed in 6.0.15.
Attempts to add a new archive path or alter a path to an archive path were being rejected by the SureSync desktop. This has been fixed in 6.0.15.
System.NullReferenceException: Unable to Clone Relation when attempting to import a Relation using the Import Relation function. This has been fixed in 6.0.15.

File Synchronization

The handle count for a SyncFiles.exe process could grow over time due to a failure releasing all the handles. This has been fixed in 6.0.15.
New code has been added to the multi-threading engine to limit the number of threads in use when the total CPU activity on the machine is high. This has been fixed in 6.0.15.
“Action suppressed. Files sizes are different with the same date.” errors could incorrectly occur when using a Multi-Mirror under a Real-Time Monitor with Advanced Open File Support enabled. This has been fixed in 6.0.15.
When processing folder operating under certain timing conditions, search handles could be left on the folder being processed which could prevent renaming the folder and could prevent scanning the folder for changes. This has been fixed in 6.0.15.
If you process a sparse file and select the rule option to verify the CRC on all files, you would receive an error “verify of write failed for destination”. This has been fixed in 6.0.15.

Real-Time Monitors

The Change Journal could be initialized more than once when a Real-Time Monitor was started. This looks to be harmless but was a waste of resources. This has been fixed in 6.0.15.
“System.ArgumentException: An Entry with the same key already exists” errors could occur in a Real-Time Monitor which would place that monitor on Hold. This has been fixed in 6.0.15.
In a Real-Time Monitor, scans could be triggered unnecessarily if a journal checkpoint was taken for an event that is normally not processed. This has been fixed in 6.0.15.
If a Real-Time Monitor is idle and no new events are being presented, a file or open retry may be delayed longer than it should be. This has been fixed in 6.0.15.
If a rename fails, perhaps because of multiple renames in a row, the “new name” may not be processed. This may mean the file is not copied until a rescan is performed. This has been fixed in 6.0.15.
If duplicate events are received for a file that will not be processed by SureSync, it is possible for events to be processed out of sequence. Optimization could discard events that should be processed in this situation. This has been fixed in 6.0.15.

SPIAgent

The SPIAgent could encounter a time out when waiting for a session mutex during the “Remove SPIAgent ID” function. This would lead to path drops and the appearance of the of job hanging. This has been fixed in 6.0.15.
The SPIAgent could encounter a SPIAgent Return Code=ATimeOut and then the job would appear to hang. This has been fixed in 6.0.15.

SPIAgent File Locking

In rare circumstances, a driver timeout could occur when using file locking which would can server performance to be hampered until the SyncFiles.exe was ended. This has been fixed in 6.0.15.

SyncCmd

New operands have been added to SyncCmd. /A”Server” will start all SureSync services on the named server. /B”Server” will stop all SureSync services on the named server. If no server name is provided then the current machine will be assumed. This can be combined with other operands such as /XStopAll or /XStartAll. This has been fixed in 6.0.15.

Please note that 6.0.15 includes a new version of the SPIAgent File Locking Add-on file filter driver. This will require a reboot of each machine when you’re applying the upgrade if your machines are running with the file locking Add-on.

We recommend performing a full rescan on each Real-Time Monitor in your environment after upgrading to 6.0.15. This can be accomplished by right clicking on the Real-Time Monitor and selecting “Re-scan files”.

Return to top of page

SureSync 6.0.13 Known Issues

Version 6.0.13 was posted on June 11, 2010.

File Synchronization

When using multi-directional Rules, a “changed in multiple locations” warning could be generated erroneously. This occurred only when a directory scan was broken into pieces due to performance of the scan or file locking being enabled. This has been fixed in 6.0.14.
In some circumstances, the calculation of a root folder scan time was skipped which caused incorrect values to be used when determining if a folder scan would be more efficient than reading change journal entries. This could result in extra folder scans occurring. This has been fixed in 6.0.14.
Extraneous error messages could be reported when Advanced Open File Support is enabled. In a monitor, with Advanced Open File Support turned on, a file copy could be performed successfully and an error still reported in the job log indicating that the file was not copied. This would be followed by the correct message indicating a successful copy. This has been fixed in 6.0.14.
With multi-directional Rules, the file history was not always deleted from the database properly for a folder that was deleted. This has been fixed in 6.0.14.

SPIAgent File Locking

When using SPIAgent File Locking, a Real-Time Monitor could encounter a deadlock situation when attempting to release a lock. When this occurs, the monitor would become unresponsive and would have to be stopped with “End Process”. This has been fixed in 6.0.14.
When using SPIAgent File Locking, if multiple users locked the same file that allows multiple user access, such as an Access database, the locks were not properly reported in status. This has been fixed in 6.0.14.
When using SPIAgent File Locking, file and folder deletes were generating locks unnecessarily. These locks were not always released properly. Deletes will no longer generate locks. This has been fixed in 6.0.14.
When using SPIAgent File Locking, file and folder renames could trigger a delete and add instead of a rename. This has been fixed in 6.0.14.
When using SPIAgent File Locking, cases could occur where a lock would not be released after a file closed. This has been fixed in 6.0.14.

Please note that 6.0.14 includes a new version of the SPIAgent File Locking Add-on file filter driver. This will require a reboot of each machine when you’re applying the upgrade if your machines are running with the file locking Add-on.

Return to top of page

SureSync 6.0.12 Known Issues

Version 6.0.12 was posted on June 3, 2010.

File Synchronization

When a rescan of an individual Relation with in a Real-Time Monitor is requested, the Relation may not be scanned. Scanning the entire Monitor or from within a Schedule worked properly. This has been fixed in 6.0.13.
In certain timing situations, a Schedule could hang when a Relation ends while retries are pending. The user could still cancel the Relation and the Schedule would continue. This has been fixed in 6.0.13.
If an individual job had thousands of records in the job log, the database could timeout while attempting to purge out the records when they become obsolete. This has been fixed in 6.0.13.
The synchronization engine could loop and consume 100% of the CPU during certain file retry operations until the retry succeeds. This issue is more likely to be common when file locking is used. This has been fixed in 6.0.13.

Installer

The installer was failing to install the SPIAgent File Locking components on Windows 2008 R2 machines due to a change in the installer application used to build the installer. This problem was introduced in 6.0.12. This has been fixed in 6.0.13.

Real-Time Monitors

In certain circumstances, a Real-Time Monitor could encounter an “Error: System.Data.SqlClient.SqlException: Violation of PRIMARY KEY constraint ‘PK_FileHistory_Rel_FileId_RootNumber’. Cannot insert duplicate key in object ‘dbo.FileHistory’.” message which would cancel the Monitor and place it on hold. This has been fixed in 6.0.13.

SPIAgent

A SPIAgent service could crash when handling a large number of sessions at the same time. This has been fixed in 6.0.13.
“Logon as batch job” rights were being enforced for connecting to a SPIAgent. While “logon as batch job” rights are recommended for the logon accessing an Agent, things will work OK without it. The user would see a “You have successfully communicated with the SPIAgent service on [machine name] but your login credentials failed” error on a Test in SPIAgent Setup and SPIAgent Return Code=”AAPIError” messages trying to browse to paths on the SPIAgent machine. This has been fixed in 6.0.13.

User Interface

Enhancements have been made to the Detailed Status panel. Locks are now displayed in a separate pane of the display, the display can be frozen to allow rapidly changing items to be examined, and new filter capabilities have been added. This has been added in 6.0.13.

Return to top of page

SureSync 6.0.11 Known Issues

Version 6.0.11 was posted on May 7, 2010.

Desktop

Machines configured via a registry setting or a group policy object to prevent the use of non-FTPS certified algorithms for cryptography would prevent SureSync’s desktop from loading. The algorithm used for database encryption has been changed to be FTPS certified. If using an existing database, you will have to install SureSync on a machine which allows non-FIPS compliant algorithms and perform a Copy Database to convert the encryption. This has been fixed in 6.0.12.

Enterprise Status

If you attempt to specify a non-default TCP listen port to connect to the remote Scheduler on in Enterprise Status an error would be generated. This has been fixed in 6.0.12.

File Synchronization

In some circumstances, a “Failed to enable constraints” error message could occur in the queue table of the database. This has been fixed in 6.0.12.
Synchronization jobs could generate “SqlTypeException: SqlDateTime overflow” errors which would cause the job to cancel. This has been fixed in 6.0.12.
If a path dropped from a synchronization job during a folder scan, the job could hang. This has been fixed in 6.0.12.
Corrections have been made to the network throttling code to enhance accuracy of the throttle. This has been fixed in 6.0.12.
The ability to filter based on additional attributes has been added to the product. You can now filter off of the offline, symbolic link and indexed attributes. This has been added in 6.0.12.
A new internal I/O throttle has been added to the multi-threaded engine to slow down I/O requests to a destination machine which is displaying reduced I/O subsystem performance. This has been fixed in 6.0.12.
A new enhancement has been made to the processing of a full directory scan. When performing a scan over slow connections or when there are pending unlock requests from the SPIAgent File Locking Add-on, the monitor will process the root folder and queue scan events for each subfolder instead of processing the entire tree at once. This allows the scan to resume where it left off and process unlock requests during the scan. This has been added in 6.0.12.

Real-Time Monitors

“Missing FileHistory” errors could occur in a Real-Time Monitor following a rename. This has been fixed in 6.0.12.

SPIAgent

Internal corrections have been made to the SPIAgent to reduce CPU and resource usage. This has been fixed in 6.0.12.

SPIAgent File Locking

Locks generated during an initial scan were not being released until the scan was completed. On long running scans, this could result in files being locked which were not really open anymore. This has been fixed in 6.0.12.
In rare circumstances, a program check could occur in the Socket I/O thread of the SPIAgent that would crash the service. This has been fixed in 6.0.12.

Return to top of page

SureSync 6.0.10 Known Issues

Version 6.0.10 was posted on April 26, 2010.

FTP

A ‘File Not Found’ error was possible when using a FTP path as the source. This has been fixed in 6.0.11.

Licensing

An unnecessary warning message was displayed for a workstation license if the license was activated for the current workstation and no root paths had been referenced. This has been fixed in 6.0.11.

Real-Time Monitors

In some instances, a folder rename event could be missed in a Real-Time Monitor. This has been fixed in 6.0.11.

SPIAgent

During a full scan where security must be compared, the SPIAgent service could consume high levels of CPU time while performing the comparisons. This has been fixed in 6.0.11.

SPIAgent File Locking

A file rename on a file that is locked with the SPIAgent File Locking Add-on could generate an ANotFound error when attempting to release the lock. This has been fixed in 6.0.11.

Please note that 6.0.11 includes a new version of the SPIAgent File Locking Add-on file filter driver. This will require a reboot of each machine when you’re applying the upgrade if your machines are running with the file locking Add-on.

Return to top of page

SureSync 6.0.9 Known Issues

Version 6.0.9 was posted on March 30, 2010.

File Locking

When using the SPIAgent File Locking Add-on on Windows 2003, file locks were not always released properly. This has been fixed in 6.0.10.

File Synchronization

VSS Advanced Open File Support would not be used in certain circumstances due to an erorr in a check to determine if the file had changed since initial processing started. Generally, the file copy would succeed using the standard open file handling. This has been fixed in 6.0.10.
A failure sending an e-mail alert could result in the job being canceled. This has been fixed in 6.0.10.
“System.NullReferenceException: Object reference not set to an instance of an object” errors could occur due to a timing related error purging internal node items. This has been fixed in 6.0.10.
InvalidCastException errors could occur when synchronizing with Novell machines. This has been fixed in 6.0.10.
Open file processing via VSS would not be used if file being processed increased in size during the processing. The standard open file support would be used instead. This has been fixed in 6.0.10.

Installer

When installing the SPIAgent File Locking Add-on on a Windows 2003 server, the user could encounter an “Installation Failed” pop-up and the driver would not be installed. This has been fixed in 6.0.10.
Corrected some display errors in the SureSync setup user interface when running on a Windows 2003 machine over remote desktop. This has been fixed in 6.0.10.

Real-Time Monitors

When replacing a file in a Multi-Mirror with a copy of the file that had an older date, the Monitor would select the file with a newer modification date on another path as the source for the copy when the older dated restored copy should have been selected. This has been fixed in 6.0.10.
PRIMARY KEY constraint ‘PK_FileActions_Rel_NodeId_FileName’ errors could occur in a Real-Time Monitor which processed multiple rename events in a particular timing. This has been fixed in 6.0.10.

SPIAgent

In some instances, the Application Event log would record a large number of “CtlSignalCaller unable to restore ReqB” messages when running with the SPIAgent Add-on. This has been fixed in 6.0.10.

User Interface

When using the Job Log Viewer with SQL 2000, the user would receive a “System.Exception: ‘Row_Number’ is not a recognized function name” error. This has been fixed in 6.0.10.
Some dates in the SureSync desktop user interface were not properly localized. This has been fixed in 6.0.10.

Please note that 6.0.10 includes a new version of the SPIAgent File Locking Add-on file filter driver. This will require a reboot of each machine when you’re applying the upgrade if your machines are running with the file locking Add-on.

Return to top of page

SureSync 6.0.8 Known Issues

Version 6.0.8 was posted on March 18, 2010.

File Synchronization

A job could encounter an “Error on BackupRead. The parameter is incorrect.” error in rare circumstances. This has been fixed in 6.0.9.
A Schedule could get “stuck” when moving between Relations when a Schedule was configured to run multiple Relations. The job would have to be terminated through Task Manager. This has been fixed in 6.0.9.
If a file system reported that it supports named data streams but the BackupRead API returned a not supported message, it was possible in rare circumstances to get a zero length destination file copy. This has been fixed in 6.0.9.
When performing a folder rename in a Multi-Mirror rule, it was possible to encounter a “Missing FileActions row” error message. This has been fixed in 6.0.9.
If using Advanced Open File Support and processing a open file which changes constantly, it was possible to have that file copy not be processed in a timely manner. The file would keep getting queued for retry until the activity slowed down because the file would change between the Rule being processed and the VSS snapshot being opened. This has been fixed in 6.0.9.

Job Log

Job log messages were being time stamped with the time they were entered into the log instead of the completed time of the action. This would cause entries to appear out of order in situations such as a path drop. This has been fixed in 6.0.9.

Licensing

SureSync was allowing workstation licenses to be activated when only server licenses were present. This would cause issues activating the servers because the workstations had used those licenses. This has been fixed in 6.0.9.
Corrections have been made to the licenses panel to display workstation license information more clearly. This has been fixed in 6.0.9.

Real-Time Monitors

When a Real-Time Monitor encountered an error which should have placed the Monitor on hold, the Monitor would sometimes restart again. This could result in the Monitor looping through the error and restart process. This has been fixed in 6.0.9.
When a path was dropped from a Real-Time Monitor, the path could be immediately dropped again. Recovery would happen fully on the second retry. This has been fixed in 6.0.9.
“System.IndexOutOfRangeException: Index was outside the bounds of the array.” error messages could occur in a Real-Time Monitor which is configured to not use history. This has been fixed in 6.0.9.

SPIAgent

If you configure the local agent in SPIAgent Setup with a DNS name, SureSync would be unable to communicate with the local agent. This has been fixed in 6.0.9.

SPIAgent File Locking

If you are monitoring subfolders with a Rule in a Real-Time Monitor with file locking enabled, an “ANotFound” error could occur when attempting to release locks. This has been fixed in 6.0.9.
When creating a new folder or file using a right click with file locking enabled, it was possible for a lock to not be released. This has been fixed in 6.0.9.

Please note that 6.0.9 includes a new version of the SPIAgent File Locking Add-on file filter driver. This will require a reboot of each machine when you’re applying the upgrade.

Return to top of page

SureSync 6.0.7 Known Issues

Version 6.0.7 was posted on March 3, 2010.

File Synchronization

Advanced Open File Support through VSS was generating errors on 64-bit Windows 2003 machines. This has been fixed in 6.0.8.
When processing a job, SQL errors could be generated when encountering folder names containing a single ‘ character. This would cancel the job. This has been fixed in 6.0.8.
When using a source path that is a VMWare based Virtual Server, it was possible for the operating system to be incorrectly identified as Netware. This resulted in errors enumerating the contents of the folders. This has been fixed in 6.0.8.

Licensing

An error in the database upgrade process was clearing server/workstation indicators. This could allow someone to activate a workstation as a server, leading to a licensing problem. This has been fixed in 6.0.8.
SureSync was incorrectly requiring activation for workstation destinations when using a workstation license. This has been fixed in 6.0.8.

Real-Time Monitors

File and folder renames were not always being processed correctly, particularly when the rename was simply a case change. This has been fixed in 6.0.8.
When running a Real-Time Monitor, the job could generate “System.Data.DeletedRowInaccessibleException” error messages. This has been fixed in 6.0.8.
When running a Real-Time Monitor which uses Intervals, the Monitor could generate System.NullReferenceException when going between Intervals. This has been fixed in 6.0.8.
The SureSync desktop will now display that it is “(Between Intervals)” on the Desktop and in the Detailed Status panel. This display had been removed and has been put back. This has been fixed in 6.0.8.

SPIAgent Add-On

A job could hang when the connection a source SPIAgent was lost. This has been fixed in 6.0.8.
A couple of situations where the SPIAgent service would generate 500 repeated errors resulting in the service being stopped with a message “**** Process terminated due to too many error messages. ****” existed. This has been fixed in 6.0.8.

SPIAgent File Locking Add-On

With the SPIAgent File Locking Add-on, it was possible to encounter “SPIAgent Return code=ANotFound” errors when releasing a lock. This has been fixed in 6.0.8.

User Interface

In some circumstances, the job log viewer would not display the job log records for a Relation, Schedule or Monitor even though the log entries existed in the database. This has been fixed in 6.0.8.
When previewing a stand-alone Relation, that previewed job could be displayed in Enterprise Status causing problems with the operation of the Enterprise Status panel. This has been fixed in 6.0.8.
If a Relation had been run in only a Real-Time Monitor, the date on the Statistics tab of the Relation would report 1/1/1900 or later. This has been fixed in 6.0.8.
Using the Test button in Tools | SPIAgent Setup to test to a machine accessed via UNC path would result in an internal error occurring. This has been fixed in 6.0.8.
If the user defines a custom location for User Interface traces, this value was not always being used immediately. This has been fixed in 6.0.8.
When removing a server from the list in Enterprise Status, a SQL error could be generated and the server would not be removed. This has been fixed in 6.0.8.

Please note that 6.0.8 includes a new version of the SPIAgent File Locking Add-on file filter driver. This will require a reboot of each machine when you’re applying the upgrade.

Return to top of page

SureSync 6.0.6 Known Issues

Version 6.0.6 was posted on February 19, 2010.

Desktop

In some circumstances, the SureSync desktop would sit on “Loading Registry Settings” for minutes before proceeding. This has been fixed in 6.0.7.
When browsing the Network node in the Add a Root Path or Edit a Root Path panels, some machines in the list could be duplicated. This has been fixed in 6.0.7.
When importing a permanent license, the license could be incorrectly marked with an expire date. Please re-import your license to solve this problem. This has been fixed in 6.0.7.
If a Root Path was modified in any way with a SureSync installation without ArchiveSync licenses, the “Do Not Monitor Path” option would get turned off and the desktop would prohibit you from turning it back on. This results in Monitors not picking up changes from that path. This has been fixed in 6.0.7.

File Locking

When using the SPIAgent File Locking Add-on, it was possible to encounter an “ANotFound” error when attempting to release a lock. This has been fixed in 6.0.7.
In some circumstances, when deleting files with the SPIAgent File Locking Add-on enabled, files within the Root Path which were not actually open would be locked. This has been fixed in 6.0.7.
When adding a large number of folders and files to a Real-Time Monitor with File Locking enabled, it was possible for some locks to not be released until the Monitor is stopped. This has been fixed in 6.0.7.
When viewing a Real-Time Monitor with File Locking enabled in Detailed Status, it was possible to receive a System.InvalidOperationException: DataTable internal index is corrupted error when a large amount of locks were being displayed at once. This has been fixed in 6.0.7.

FTP

When using FTP, the connection could be timed out by the FTP server during a directory scan. Changes have been made to prevent the timeout. This has been fixed in 6.0.7.
When using FTP, status was not being reported correctly. This has been fixed in 6.0.7.

Real-Time Monitors

Made corrections to Real-Time Monitor File History when one or more paths were dropped and file changes occurred during the path drop. A full rescan of all Real-Time Monitors is recommended to ensure all files have been synchronized. This has been fixed in 6.0.7.

Return to top of page