# Strange problem with tag reads potentially caused by subscription loss

**URL:** <https://forum.inductiveautomation.com/t/strange-problem-with-tag-reads-potentially-caused-by-subscription-loss/109900>\
**Category:** Ignition\
**Tags:** ignition81\
**Created:** [August 18, 2025, 6:16pm UTC](https://forum.inductiveautomation.com/t/strange-problem-with-tag-reads-potentially-caused-by-subscription-loss/109900 "2025-08-18T18:16:32Z")\
**Posts on this page:** 9\
**Page:** 1

<div class="post-metadata">

**Author:** ![Kenneth\_Wainwright](https://avatars.discourse-cdn.com/v4/letter/k/e99b99/32.png) [@Kenneth\_Wainwright](https://forum.inductiveautomation.com/u/Kenneth_Wainwright)\
**Post date:** [August 18, 2025, 6:16pm UTC](https://forum.inductiveautomation.com/t/strange-problem-with-tag-reads-potentially-caused-by-subscription-loss/109900/1 "2025-08-18T18:16:32Z")

</div>

I have two issues that I thought were unrelated at first but now appear to be related.

The symptom: a full loss of tag communication over a gateway network, despite status appearing to be good

The cause? The loss of an OPC subscription with a Siemens S-7 PLC for some of those tags

We have two gateways. One I will call the "screen server" handles all of our connections with the HMI clients. It is connected over a gateway network to the "tag server". The tag server handles OPC connections to multiple S7-1500 PLCS (currently simulated via PLC Sim Advanced) and history for those tags.

The problem initially appeared as a complete loss of ability to read/write tags on the screen server over that gateway network. This despite the fact that status appears good on the connection itself and those tags can be accessed no problem from the tag server gateway. This is accompanied by an error in logs (see below) for every blocking read operation the screen server is attempting. A reset of either gateway resolves the issue until it happens again weeks later.

I believe the cause is a loss of the OPC subscription to a couple of these S7 PLCs. The tag server loses connection temporarily and then is unable to re-establish it; it attempted to do so for days over the weekend. I do not understand _why_ this would cause the problem, as the connections should be entirely separate. However, I found a similar problem on the forums with a physical PLC: [OpcServer exception thrown and didn't recover - Ignition - Inductive Automation Forum](https://forum.inductiveautomation.com/t/opcserver-exception-thrown-and-didnt-recover/92161). I am wondering if this is a bug that needs to be resolved.

Any help would be appreciated! This situation would be a disaster in production.

Ignition version 8.1.42

---

<div class="post-metadata">

**Author:** ![Kenneth\_Wainwright](https://avatars.discourse-cdn.com/v4/letter/k/e99b99/32.png) [@Kenneth\_Wainwright](https://forum.inductiveautomation.com/u/Kenneth_Wainwright)\
**Post date:** [August 18, 2025, 6:19pm UTC](https://forum.inductiveautomation.com/t/strange-problem-with-tag-reads-potentially-caused-by-subscription-loss/109900/2 "2025-08-18T18:19:02Z")

</div>

I suspect the problem begins here:

 ![Cause](https://us1.discourse-cdn.com/inductiveautomation/original/3X/e/e/ee9ca63b131997ddc7faf9eb80365a306c6635aa.jpeg)

This seems to create problems with tag reads from the screen server to the tag server

 ![Solution](https://us1.discourse-cdn.com/inductiveautomation/original/3X/8/a/8ae02b7406dc80c9f47a5a12f49296dbba10bc4d.jpeg)

---

<div class="post-metadata">

**Author:** ![Kevin.Herron](https://sea1.discourse-cdn.com/inductiveautomation/user_avatar/forum.inductiveautomation.com/kevin.herron/32/7259_2.png) [@Kevin.Herron](https://forum.inductiveautomation.com/u/Kevin.Herron)\
**Post date:** [August 18, 2025, 8:30pm UTC](https://forum.inductiveautomation.com/t/strange-problem-with-tag-reads-potentially-caused-by-subscription-loss/109900/3 "2025-08-18T20:30:35Z")

</div>

You'll probably need to troubleshoot that backend Ignition's connection to the problematic OPC UA server. It looks like that server stops responding and/or closes the connection. May need a Wireshark capture to see help see what's going on.

---

<div class="post-metadata">

**Author:** ![Kenneth\_Wainwright](https://avatars.discourse-cdn.com/v4/letter/k/e99b99/32.png) [@Kenneth\_Wainwright](https://forum.inductiveautomation.com/u/Kenneth_Wainwright)\
**Post date:** [August 18, 2025, 9:51pm UTC](https://forum.inductiveautomation.com/t/strange-problem-with-tag-reads-potentially-caused-by-subscription-loss/109900/4 "2025-08-18T21:51:15Z")

</div>

Thanks for the response, Kevin.

I figured the connection to the 3rd party device would need some troubleshooting. What I am more concerned about is the apparent connection to the loss of tag communication across that gateway network. What would you need to see to prove that this is the case?

---

<div class="post-metadata">

**Author:** ![Kevin.Herron](https://sea1.discourse-cdn.com/inductiveautomation/user_avatar/forum.inductiveautomation.com/kevin.herron/32/7259_2.png) [@Kevin.Herron](https://forum.inductiveautomation.com/u/Kevin.Herron)\
**Post date:** [August 18, 2025, 10:01pm UTC](https://forum.inductiveautomation.com/t/strange-problem-with-tag-reads-potentially-caused-by-subscription-loss/109900/5 "2025-08-18T22:01:19Z")

</div>

You seem to be doing blocking reads from scripting against those tags, in which case it’s not that surprising that it fails if the server isn’t responding or connected.

Maybe I’m missing something about what you’re saying here.

---

<div class="post-metadata">

**Author:** ![Kenneth\_Wainwright](https://avatars.discourse-cdn.com/v4/letter/k/e99b99/32.png) [@Kenneth\_Wainwright](https://forum.inductiveautomation.com/u/Kenneth_Wainwright)\
**Post date:** [August 18, 2025, 10:15pm UTC](https://forum.inductiveautomation.com/t/strange-problem-with-tag-reads-potentially-caused-by-subscription-loss/109900/6 "2025-08-18T22:15:53Z")

</div>

Ok, I tried simply disconnecting one of the PLCs and immediately recreated both problems. What is concerning is that I cannot browse _any_ tags on the tag server from the screen server when this is happening. Essentially, when one PLC goes dark, the entire system goes down. I can still browse tags no problem on the tag server. Is this really how a blocking tag read is supposed to work?

 ![image](https://us1.discourse-cdn.com/inductiveautomation/original/3X/0/2/021c056c2295f77ddaf0b2b08fff42615d3e3a25.png)

 ![image](https://us1.discourse-cdn.com/inductiveautomation/original/3X/3/6/36b57569bafb6a23a3537bd6feef0223b3a930ae.png)

 ![image](https://us1.discourse-cdn.com/inductiveautomation/original/3X/2/c/2cfe208d00126fbff0ba65e6bec3a55d60c25238.png)

---

<div class="post-metadata">

**Author:** ![Kevin.Herron](https://sea1.discourse-cdn.com/inductiveautomation/user_avatar/forum.inductiveautomation.com/kevin.herron/32/7259_2.png) [@Kevin.Herron](https://forum.inductiveautomation.com/u/Kevin.Herron)\
**Post date:** [August 18, 2025, 10:20pm UTC](https://forum.inductiveautomation.com/t/strange-problem-with-tag-reads-potentially-caused-by-subscription-loss/109900/7 "2025-08-18T22:20:06Z")

</div>

Call into support and let them take a look at your setup.

---

<div class="post-metadata">

**Author:** ![Kenneth\_Wainwright](https://avatars.discourse-cdn.com/v4/letter/k/e99b99/32.png) [@Kenneth\_Wainwright](https://forum.inductiveautomation.com/u/Kenneth_Wainwright)\
**Post date:** [August 18, 2025, 10:22pm UTC](https://forum.inductiveautomation.com/t/strange-problem-with-tag-reads-potentially-caused-by-subscription-loss/109900/8 "2025-08-18T22:22:26Z")

</div>

Roger that, thanks.

---

<div class="post-metadata">

**Author:** ![Chris\_Bryant](https://avatars.discourse-cdn.com/v4/letter/c/d6d6ee/32.png) [@Chris\_Bryant](https://forum.inductiveautomation.com/u/Chris_Bryant)\
**Post date:** [February 3, 2026, 8:12pm UTC](https://forum.inductiveautomation.com/t/strange-problem-with-tag-reads-potentially-caused-by-subscription-loss/109900/9 "2026-02-03T20:12:58Z")

</div>

Ken we are having a similar issue at one of our sites. What was the end result/diagnosis and issue resolution for what you were seeing? The behavior is almost identical.
