deviceWISE M2M 13.1 Release Notes
What's new in the deviceWISE 13.1 release
Supported platforms
deviceWISE M2M Enterprise Gateway products
- Software Products
- Linux (CentOS5 and CentOS6)
- Windows (Windows XP, Windows 7, and Windows Server 2008R2)
- *New* AIX (Version 6.1 and Version 7.1)
- Embedded Products
- Mitsubishi WinCPU
- Siemens ERPC
deviceWISE M2M Asset Gateway products
- Software Products
- Linux (CentOS5 and CentOS6)
- Windows (Windows XP and Windows 7)
- Embedded Products
- CalAmp LMU-5000 (Rev C and Rev G)
- Multitech OCG-D (H4, H5, and EV2)
Driver enhancements
Refer to the Resources tab on the Management Portal to identify the drivers and driver versions supported on a specific platform with this release. The following drivers were added or updated with this release.
- OPC-DA Client driver supports servers that meet the 2.0x specification.
- Added OPC-DA Server support
- Added Modbus RTU Server and Modbus TCP Server support
- Added Atlas-Copco Driver support
- Added ROC Plus serial driver
New features
- CloudLINK Alias
- Web Services SOAP Transport
- License management via the Management Portal
- License management via an API
Known issues for the deviceWISE 13.1 release
Enterprise Gateway for Windows
MAP-3455: The Enterprise Gateway for Windows Restart feature from Node Administration tab does not work as designed
Description: When using the Workbench in conjunction with an Enterprise Gateway for Windows node, the Restart feature available from Administration -> Node Administration tab may not cleanly stop and then restart the runtime. When there are many active transports and listeners, the Transaction Server does not restart properly.
Workaround: Follow the instructions in the section called "Restarting the deviceWISE node" on Enabling MySQL support.
Enterprise Gateway for ERPC
MAP-1933: ERPC: Controlled Listener fails to connect to queue manager on first attempt
Description: When accessing a Websphere MQ server via a transport or a listener the very first connect request after boot fails due to a limitation in the firmware. Subsequent connect requests work correctly.
Workaround: Repeat the connect request.
MAP-3634: FTP action with Transfer Mode set to Active fails on Enterprise Gateway for Siemens ERPC
Description: When using a trigger FTP action, the Put and Get operations do not work when Transfer Mode set to Active on ERPC.
Workaround: Set the Transfer Mode parameter to Passive instead of Active until the problem is fixed in a future release.
Runtime
MAP-4002: Updating logical variables in a transport map and not editing the corresponding trigger can lead to undesirable consequences.
Description: In this scenario the trigger will be disabled when it fires, and the action will return an error code of -6409 (invalid variable type) in the exception log. Because the trigger will not be recovered on it's own, the user will have to edit and re-save the trigger manually.
Transaction server
MAP-685: deviceWISE messaging transports do not recover properly from Store and Forward (FIFO and non-FIFO)
Description: When using the WMQ transport and the
transport reports the error MQJMS007, it might cause the
transport to switch in and out of Store and Forward, even
though the transactions are successfully delivered to the
WMQ queue.
Check the WMQ queue manager. If the listener shows the
state other than Started, then it is
likely that you have encountered this issue. Set the
listener to Started to restore proper
functionality.
MAP-2972: Siemens ERPC node faults with Transaction Engine not being able to connect after package remove
Description: After an install or removal of a deviceWISE driver package, a Siemens ERPC node might have a fault where the transaction server cannot connect.
Workaround: You will need to restart deviceWISE. Follow the instructions for Manually restarting a deviceWISE node on Packages.
MAP-3179: Listener does not process the first delimited request that comes in
Description: The TCP transport may not detect the connection to the host has been lost and re-established. As a result, if the TCP transport was connected and its state "UP", and the host reset the connection, the first transaction sent through the transport after the connection has been re-established will fail. For more information, see TCP listener might not receive first message.
MAP-3507: Select with Update fails to update the TIMESTAMP with TIME ZONE column values
Description: There are two limitations:
- Stored procedures with INOUT parameter of type TIMESTAMP with TIME ZONE are not supported.
- Select with Update of a TIMESTAMP with TIME ZONE parameter will get an SQL exception and is not supported. The Select with Update fails to update the TIMESTAMP with TIME ZONE column values. The select completes successfully, but the update fails on only the TIMESTAMP with TIME ZONE type. All other column values are updated successfully.
For more information, see Oracle database with time stamp with time zone.
MAP-3615: TCP transport needs to detect the connection has been terminated
Description: When using a TCP transport to "feed" a TCP
listener and the listener is stopped and restarted, the
transport does not "notice" the connection has been
dropped. The next transaction appears to be sent
successfully and the transport success count incremented by
one. Yet the transaction was not received by the listener.
The second transaction sent by the transport will get a
"connection reset by peer" error, enters its retry logic
and this second transaction is successfully processed by
the listener.
However, a message was lost and no warning to the user
since the transaction was marked as successful. For more
information see the Limitations section on
Defining a TCP transport.
MAP-3930: IFS: sometimes transaction server fails to start
Description: When setting transports, especially an iTAC
transport, with "Load transport at initialization"
parameter checked, and using a memory optimized Transaction
Server is not a match for the transports with "Load
transport at initialization", it is possible that the
Transaction Server will fail to start on Mitsubishi
MESInterface IT nodes.
Rebooting the module with the "Standard" Enterprise Access
option will allow the Transaction Server to come up
successfully, the transports can then be modified to remove
the "Load transport at initialization" parameter. The
memory optimized Transaction Server can then be re-set to
the desired one.
Transaction server XSD support
MAP-3141: XSD: Xpath expression should not be saved in XML if Specify Map and Sequence Keys in header check box is selected
Description: This issue only affects JMS listeners with the XSD payload. When there is both an XPath expression and a JMS header property specified on the JMS listener XSD Payload panel, and the header property is missing from the incoming JMS message, the XPath expression is used to determine the listener map id, instead of the default listener map id.
MAP-3695: XSD: Default or fixed value provided in the XSD will be erroneously used
Description: If the XSD used to create a listener map contains a required attribute with a default or fixed value, it is considered valid. At runtime, instead of the XML being rejected as invalid, the default or fixed value provided in the XSD will be erroneously used.
MAP-4665: XSD: Imported transport/listener map definitions from an older runtime into a newer runtime are not namespace aware unless xsd is requeried
Description: If you are importing a transport or listener map that was defined with XSDs and was created in a 12.4 or prior environment, then the map needs to queried again and saved in order for 'nil' element processing to be handled correctly.
MAP-4666: Listener doesn't support XSDs with namespaces
Description: A Listener configured to process XSD based payloads will not be able to successfully handle an XML with namespace qualified elements.
Workaround: The workaround is to provide an XML request without namespaces.
Transaction Server Web Services SOAP
MAP-4670: WebServices transports WS_SOAP, IFS and SAP_MII do not behave correctly when Suspend is called.
Description: After you Suspend a SAP_MII, WS_SOAP or IFS transport, transactions continue to be delivered to the target enterprise application instead of failing or going to Store and Forward.
Workaround: The workaround is to change the transport host name or port settings to a value that will prevent the transport from successfully establishing a connection to the enterprise application.
Drivers
Modbus Server
MAP-4664: Modbus Server doesn't remove lost connections on Linux
Description: Linux does not detect when a client connection has been unplugged, so the server maintains the connection. The server code is written to attempt a timed read from a connected socket. The amount of time it waits on the read is set by the user in the server's Timeout field. The read will fail with a timeout error if there is no activity on the line. In situations where a read was successful or if the read failed due to a timeout, the code circles back and attempts another timed read. This continues until the server is stopped or if the client closes the socket connection.
Windows will recognize the disconnect state due to a client losing the Ethernet connection. The server's read will result in a non-timeout read error, allowing the server to close the connection.
For more information, see the section "The client's Ethernet/TCP connection to a Linux Modbus Server won't go away" on Modbus Server troubleshooting.
OPC-DA Client
MAP-3598: Rockwell RSLinx OPC-DA Server is not supported
Description: The deviceWISE OPC-DA client device driver does not support the Rockwell RSLinx OPC-DA Server. For more information, see Supported OPC-DA Environments.
Siemens S7 listener
MAP-3140: Siemens S7 listener gets stuck in stopping state on Enterprise Gateway for Siemens ERPC
Description: After the Siemens S7 listener receives data and a corresponding trigger is executed, stopping the Siemens S7 listener results in the listener becoming stuck in the Stopping state. This happens during operations when the Siemens S7 listener is using the normal TCP mode (not ISO-n-TCP). After a trigger executes the receipt of an unsolicited message, the S7 listener cannot be stopped.
A firmware upgrade from Siemens is anticipated in order to resolve this behavior.
MAP-3470: Siemens S7 native driver loses variable registers when more than two devices are started simultaneously
Description: When three or more deviceWISE devices that access the same Siemens PLC are started simultaneously, the Data Block (DB) registers might enumerate correctly or they might not enumerate at all.
To correct this problem, simply restart the deviceWISE device that exhibits the enumeration problem. To prevent this problem from occurring, stagger the start of the devices so they do not all start concurrently.
Workaround: A delay between 50 and 100 milliseconds between the start of each device should be adequate to ensure the Data Block registers enumerate correctly for all devices.
Modem
MAP-3167: Simultaneous Cellular connect/disconnect and Modem power on/off can cause reboot of Asset Gateway
Description: We found a scenario where the Asset Gateway will reboot when simultaneous control actions are executed. When the Set Modem Power Level action and the Control Cellular Connection action (or Workbench operation) is executed repeatedly and simultaneously, a crash can occur resulting in a reboot.
We have not seen this problem during normal operation where the use of the Set Modem Power Level action cycling of power between "Online" and "Low Power" rarely occurs.
Workbench
MAP-4588: Workbench cannot reconnect when CalAmp LMU-5000 Rev G node is scanned over CloudLINK after package add or remove
Description: When taking an action on a CalAmp LMU
(Revision G) that requires the node to restart - e.g.
installing or removing a package, or manual restart - the
node may become unresponsive to workbench commands.
This can be corrected by either rescanning the node (from
the "Nodes" entry in the navigation tree), or exiting and
restarting the workbench.
Fixed defects in deviceWISE 13.1 release
The following defects were highlighted as known issues in previous releases, and were fixed in this release.
Transaction Server
MAP-4006: Null pointer error message from Transaction Server
Description: When a database transport is non FIFO and the store and forward file is full, there is a small possibility that one transaction will be marked as failure and a severe error noted in the log when the transport is first able to connect to the host.
MAP-4007: Transaction Server - transaction stuck in Store and Forward queue
Description: When a database transport has Store and Forward set to FIFO (process queue before leaving Store and Forward) it is possible that after all transactions have been delivered to the host and the transport has switched to "UP", that the counter still shows one transaction in the Store and Forward queue.
MAP-4296: Transports in non-FIFO mode can leave orphaned pool threads forever when they have transitioned in and out of S&F multiple times
Description: With database transports the following combination of 'Process queue before leaving S&F' unchecked and 'Load Transport at Initialization' checked and working with a link to the database that is subject to frequent disruptions can result in a growth in the connections to the database server. Under some conditions this can also lead to the corruption of the Store and Forward file and a loss of stored data. The workaround is to not use this combination of settings.
Drivers
OPC-DA Client
MAP-3848: The OPC DA client does not support servers that meet the 2.0x spec.
Description: The OPC-DA client will only interact with OPC-DA servers that have implemented the 3.0 version of the OPC Specification. For more information see Supported OPC-DA Environments.
Upgrading from previous versions
Backup before upgrading!
You should backup your configuration from the Workbench
prior to upgrading to a new version of deviceWISE.
For more information about upgrading your software, see
Updating products.