Sign in to follow this  
Followers 0
Powered By Dodge V8

RSLinx Classic Gateway OPC Server Bottlenecks / Fails on ControlLogix 5000 Presence Detection

13 posts in this topic

Hello, Not often one to ask for help, but here's a real kicker. RSLinx Classic Gateway (aka Professional) OPC Server version 2.57 (latest I believe), running on Windows Server 2003 R2 Service Pack 2 32-bit (also had this on Windows XP Service Pack 3 32-bit and it behaved the same way). Server has 4 GB memory DDR 1066 memory / 4 Intel Xeon 2.53 GHz cores / and SAS 15k RPM hard drives in Raid Mirror. We do a lot of OPC read / write work over DA-2, and for everything except ControlLogix CPU's, this is not an issue. The first half of the problem occurred when a ControlLogix CPU was powered down... RSLinx slowed down to a crawl, would not create new groups (reads) or perform writes while it was eating up CPU apparently trying to make sense of the orphaned ControlLogix CPU. The second half of the problem occurs with the latest generation Logix CPU we have. RSLinx OPC Server seems to choke every so often with an error along the lines of "failed to read @ispresent from controller at..." or "timed out reading @ispresent from controller at..." When this happens, we get a general slow down (its a massive performance hit but because of the size of the server it doesn't actually crash it). With any other PLC (MicroLogix, SLC, PLC-5), if we are trying to create an OPC Group to read from (or even perform a write to) a PLC that is powered down, unplugged, or otherwise just not connected at the moment, then RSLinx will behave properly -- for example: -- in the case of a write, it will attempt to perform the write, but return the quality of the write as "bad" or "failed" (I forget the exact attribute terminology). -- in the case of building a group for a read, it will build the group, return data for any PLC's in that group which are UP / Online (quality attribute of data is labeled as 'Good') and return either zero's or "Error" or "Undefined" (or something like that) for any PLC's in that group which are DOWN / Offline (quality attribute of data is labeled 'Bad' or 'Failed' or whatever). With a ControlLogix 5000 series (and specifically this newest one with firmware version 19), RSLinx is choking. It can't have anything to do with the controllers - because (now think about this kids) if it had something to do with the PLC itself, then the problem would go away when the PLC was removed from the network / powered down, and RSLinx would go back to behaving normally. Any help would be appreciated, this is currently crippling my infrastructure, as we've began to deploy more and more hardware that relies upon this server lately, and this bottleneck is becoming a game-stopper.

Share this post


Link to post
Share on other sites
RSLinx Classic 2.57 is almost six years old; I'd strongly recommend updating to at least 2.59.02. CLX Firmware revision 19 is three years old, however, it has introduced plenty of "features" which most likely conflict with your "mature" RSLinx data server. You could download the RSLinx Lite 2.59.02 at http://compatibility.rockwellautomation.com/Pages/MultiProductFindDownloads.aspx?crumb=112&toggleState=&versions=51543 and use the same Gateway activation.

Share this post


Link to post
Share on other sites
I apologize for the mis-information ... just checked the help tab and we are running 2.59.00 CPR 9 SR 5 ... 2.57 is what we had launched the system on a few years back, but it's been kept up to date since then, hence the current 2.59.00

Share this post


Link to post
Share on other sites
Have you been monitoring the RSLinx CIP Diagnostics Tool (Communications/CIP Diagnostics...) when you are experiencing the described issues? How many Logix Topics is RSLinx Gateway servicing? Could you post is Screen shot of the RSLinx/Communications/Configure CIP Options.../Connections for ControLogix Processor...? Have you considered using RSLinx Enterprise as your data server: RSLinx Classic is, just as its name suggests, a legacy communications software; you might've just pushed it to its limits.

Share this post


Link to post
Share on other sites
-- less than 5. The vast majority of the facility is PLC-5 and SLC 5XX series (either DH+ via CLX Gateway or 505's over ethernet) or Micrologix. Edited by Powered By Dodge V8

Share this post


Link to post
Share on other sites
RSLinx Enterprise is an OPC data server developed for and "native" to Logix control platforms; it is used in conjunction with the FactoryTalk SCADA platform and optimized specifically for Logix class CPUs communications implementations. I looked over your RSLinx Gateway setup and trying to acquire the big picture. How do you access the Logix CPus? Are there dedicated 1756-ENB(2)T bridges on each Local Chassis patched to the "supervisory" LAN RSLinx is deployed onto or you are routing communications via backplanes and/or I/O or HMI "traffic" subnets? Are you using managed switching? What type of RSLinx driver are you using for the configured Logix topics? Ethernet or EtherNet/IP?

Share this post


Link to post
Share on other sites
-- We're using the ETH driver for everything (the one where you explicitly declare the IP address and optionally the connection type, for example 10.1.1.99 or 10.1.1.99:EIP). The primary ETH driver has about 150 PLC's on it. The second ETH driver has the two DH+ Gateways on it. And the 3rd ETH driver just has out HMI's (panelview plus) to allow remote programming.

Share this post


Link to post
Share on other sites
I "see" some pretty serious issues with your topology. I am aware that this was built up over time: initially you were "pulling" data off legacy CPUs which were happily talking to the Wonderware SCADA using RSLinx Gateway over DH+. Then, within the same topology, you started adding Logix class CPUs following the same networking approach and the same protocols due to familiarity, convenience and already existing infrastructure. For a while everything seemed to be fine, however, when the Logix firmware started to be updated, you started introducing obvious bottlenecks since, again, RSLinx Classic was and still is not optimized for deploying onto Logix systems; moreover, you are using a "plain Ethernet" driver and not an EtherNet/IP one; the RSLinx Ethernet driver uses PCCC encapsulation and it is to be deployed only within legacy firmware (Ethernet PLC5, SLC5/05); it will "work" within a Logix platform, however, returning the "results" you have been experiencing. I don't quite know what to tell you about mitigation; I'd start by deleting the Ethernet drivers used to access the Logix Topics and reconfigure EtherNet/IP ones: then I would attempt to segregate the DHRIO arriving legacy controllers data from the one generated by the CLXs; I don't know if this is possible without physically modifying the existing topology, however, this should be your main goal. You will eventually have to completely separate the legacy and Logix platforms using at least two different data servers "connected' to the same SCADA; I don't know if this is quite achievable, however, the existing setup is not viable anymore especially if intended to be used indeterminately.

Share this post


Link to post
Share on other sites
For the moment this issue has been taken out of my hands - there is a belief among some that it is a hardware issue - which is mind boggling when you consider that it occurs when the device in question is --OFFLINE--. Will provide some feedback if/when this ever gets resolved.

Share this post


Link to post
Share on other sites
just a bump here, and an update. about 2 months after it cropped up, it transitioned back to me - such is life. implementing a plan based on your thoughts as well as some other internet-dwelling psychopaths, i do believe it may be resolved. -- micrologix and slc controllers are tied in via AB-ETH driver -- devices (slc / plc-5 / other) that are accessed via a controllogix chassis bridge (just a ENBT/EN2T card and a DHRIO card in a common chassis) are accessed via the AB-ETH/IP driver -- controllogix processors are accessed via the AB-ETH/IP driver @ispresent not found no longer presents itself. rather, when the controllogix in question, or any other one for that matter, goes offline - rslinx responds with the (appropriate) error "communication lost to device [whatever device]". when the device is restored, communicate resumes shortly thereafter. we'll see how this plays out - but i'm unofficially declaring "tsang tsung - flawless victory" on this one.

Share this post


Link to post
Share on other sites
Keep your fingers crossed... From my personal experience, seamlessly integrating Logix and Legacy platforms is a myth, even if Rockwell Automation claims the opposite...After all, CLXs and PLC/SLCs truly are apples and oranges hence SCADA integration will have to reflect the "fruit differences"... Edited by dmargineau

Share this post


Link to post
Share on other sites
Experienced similar issue in the past with version 2.53, offline PLCs will cause RsLinx to slow down and eventually crash. Issue was solved by implementing a heartbeat counter in every PLC and by taking items out of scan when such heartbeat was not detected. We had Rockwell on site, they acknowledge this issue was by design. ap

Share this post


Link to post
Share on other sites
Hi All, I have the same issue, but wit RSLinx 2.59.22, which is the latest revision. Using ControlLogix 5300, firmware revision 20. RSLinx driver is Ethernet Device ETH-1 1756. Scenario: The system has 10 AB PLCs connected to different level 1 rockell switches. RSLogix is installed on redundant Servers 2008 R2. Once any one PLC or more go offline, it will crash the the remaining OPC communication with remaining PLCs. So losing any PLC will cause carshing of the remaining OPC! Any suggestion?

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0