1769-l35e

lostcontrol

Lifetime Supporting Member
Join Date
May 2009
Location
NeverSayNever
Posts
1,064
Hi,
Have seen a couple of occurrences of this error during search:
failed to go online with controller.
Unable to open connection -- all connections may be in use


From knowledgebase, it mentions that the 100% utilization is related to continuous task config.
If I change the task to periodic, will that assist in alleviating this issue.?
I am aware that the ethernet CPU is very loaded often & have plans in place to rectify, but interim is we need to keep it going.

Is there a way to reset port remotely, so at least IO Servers can reconnect succesfully & perhaps RS5k..?


Thanks
lc
 
This refers to CIP connections. The L35E I believe is limited to 32. CIP connections will be anything that communicates to the controller via remote IO, HMI/SCADA, PLC-PLC messages.

When it comes to remote IO, you have to be mindful of how many analog IO cards are in the remote rack. They consume 1 CIP connection each. My experience is that L35e using remote IO and lots of analog cards you'll hit your CIP limit pretty quickly.

More details of what the L35e is communicating to would be handy. Also, you can log onto the L35e webpage for diagnostics to see how many CIP connections are being used at any given time. Check the RPI to those IO adaptors as well.
 
Last edited:
Thanks for replies :site:
Please check ethernet connections timing, probably they are too fast. Try to slow them down.
There is 6x VSD's & 1x Point IO rack, as well as some Produce/Consume connections to 2x other PLC's.
The VSD's are 250ms, Point IO is 100.

This refers to CIP connections. The L35E I believe is limited to 32. CIP connections will be anything that communicates to the controller via remote IO, HMI/SCADA, PLC-PLC messages.

When it comes to remote IO, you have to be mindful of how many analog IO cards are in the remote rack. They consume 1 CIP connection each. My experience is that L35e using remote IO and lots of analog cards you'll hit your CIP limit pretty quickly.
This is interesting about the analogs. There are 2x cards in the point rack.

More details of what the L35e is communicating to would be handy. Also, you can log onto the L35e webpage for diagnostics to see how many CIP connections are being used at any given time. Check the RPI to those IO adaptors as well.
As above, as well as 2x Magelis HMI's, the SCADA IO Server & a large screen for general status ( so another IO Server..)
From the webpage, I noticed that there were 2x IO servers enabled. Disabling them both allowed access.
The other telling thing was that the large screen still had 4x connections. I'm putting this down to restarts of the application & it not releasing the connection properly.

You could try adjusting the System Overhead Time Slice setting to allocate more CPU time to the backend tasks such as comms. etc. I stumbled over a great article here that does a good job explaining.

https://************/system-overhead-time-slice/
Cheers, slice is already at 50%..
Ethernet (?) CPU is consistently high ( > 70%)
Would changing this continuous task to periodic improve things?

The key things that were identified from the web server was the 2x IO Serves (they should of been redundant), as well as the large screen not disconnecting correctly.

But all things aside, getting online to the PLC should always be possible.
Is there a way to monitor the connection status & possibly release..?
I'm thinking a GSV/SSV possibly, but it might be limited to the port rather than a single connection?
 
But all things aside, getting online to the PLC should always be possible.

Not if all the CIP connections are used so it may not be possible.

You really need to understand what is using CIP connections and how many. For example your VSDs. IIRC, powerflex com cards used to use 4 CIP connections, I believe newer firmware condensed it down to 1 but for arguments sake say it's 4 since you haven't mentioned specifics.

Your VSDs alone would be consuming 24/32 connections.

Your point IO rack will consume 1 if "rack optimized" for discrete IO plus the 2 analog cards. So you are now consuming 27/32.

Your HMI/SCADA and produced/consumed tags could easily consume the remaining 5 and you're out-of-luck connecting Logix5000 to the PLC.

You need to do your homework on this one, check your firmware versions and CIP connections. Look at your HMI software and IO server configurations. If they are not properly releasing CIP connections you may have to patch/update that software. Check your produced/consumed tags, it might be better to pack comms into a MSG instruction.

L35Es are nice little units however 2 CIP connection limit is a major bottleneck.
 

Similar Topics

Hello, We have had a machine running for the last 10 years. It has 2 x Kinetix 300 drives. Around a year ago, one of the drives failed. We were...
Replies
1
Views
408
Hi, I recently tried to change the IP address of my 1769-L35E Series A, since doing this the MS indicator is flashing red and the NS indicator is...
Replies
3
Views
922
Good morning guys. Im triying to communicate 4 JXC91 with a 1769-L35E AB CPU. I already test the JXC91 with a newer AB CPU using eds files and...
Replies
2
Views
1,403
Good Morning , I'm looking for a cheap 1769-L35E processor. I know this CPU is obsolete. I'm preparing for a obsolescence project , but the...
Replies
2
Views
1,167
Hi Guys, Question - does the subject title AB processor act as a processor AND comms module? It does have communications ports available on the...
Replies
2
Views
945
Back
Top Bottom