Having trouble keeping Ignition Edge 8.3.1 in sync with Ignition Standard 8.3.1.
Both installations are new, and while most functionality on both sides is working well, the tag history sync keeps failing after about a week of uptime. I’m wondering if anyone has managed to keep this running reliably for more than a few weeks?
In my setup, the Edge gateway has been online for about a month. The Remote Tag Provider has been solid the entire time, and the initial sync with the Standard gateway worked fine at first. However, after about a week, synchronization stopped without any apparent cause.
Following Inductive Automation Support’s guidance, I’ve:
- Recreated the Gateway Network connection (GAN)
- Restarted both servers multiple times
- Adjusted Security Zones on both ends
- Reinstalled the Historian Core module
None of these steps restored reliable syncing. Eventually, I deleted all historical data and reinstalled Ignition Edge. That resolved the issue temporarily—again for about a week—before the sync stopped once more.
Ignition Standard is using MS SQL for historical storage, with roughly 300 tags being historized, so it’s not a large dataset. I’ve been waiting on further feedback from Support for a few days, but in the meantime, I’d really appreciate hearing if anyone else has run into this issue and found a reliable fix or workaround.
Inductive Automation Support has come through with a fix for this!
It seems there was an underlying issue with the Edge Historian Sync in 8.3 and maybe even 8.1 which caused a backlog of records that could not keep up and fell further and further behind.
The fix was to add an index to the Edge Historian database:
CREATE INDEX tagid_ndx ON tagdetails(tagid)
6 Likes
Thanks for the workaround. Could you clarify where exactly the CREATE query should be executed on Edge?
This may have been fixed in a 8.3.2 as there are was a fix in the release notes that mentioned Edge Remote History Sync. Maybe check that first.
The is the fix that worked for me on the Edge gateway:
- Download and run DB Browser for SQLite (in admin mode).
- Open file data\var\com.inductiveautomation.historian\internalhistorian\edge_historian.idb
- Then create the index on the tagid_ndx table.
3 Likes
Hello,
Coincidenally, I also came across these issue at a customer of us a few days ago, after we upgraded to 8.3.2. So its, in fact, not fixed in this new version.
I’ve applied David’s solution yesterday, and it all seems to be working fine now.
After adding the index, 3/4 Edge Gateways pushed their buffer to the central gateway and all data was succesfully stored in the SQL Server historian without missing a gap.
However, 1 Edge Gateway did not. I fixed this by stopping the Edge Gateway, deleting the .idb-file and restarting the gateway. After this, these tags got historized again. I suppose the buffer within the Edge Historian was too big (this gateway was the first to stop syncing) to be pushed to the central gateway.
Note that in this case, your data is lost!
Also, you do not need to restart the Edge Gateway to ensure it starts syncing its data.
Just wanted to chime in and confirm that this issue was addressed in 8.3.3:
Improved Edge Historian sync performance by optimizing tag definition queries that were causing expensive temporary B-Tree operations on large datasets.
If you encounter any other Remote Sync issues on 8.3.3+, please let us know. Thanks!
2 Likes
Using the ignition 8.3.3 docker image for Ignition Edge. We’ve noticed that, upon a fresh compose, the edge sync will work fine (it must be disabled and reenabled from the GW page each time to get it to sync, for some reason). At some point, (say 15-20 minutes) the gateway becomes completely unstable, crashing about every 2 minutes from this point forward, to which it does an automatic GW restart before crashing again after 2 minutes. This does not happen if edge sync is not currently enabled.
Could you upload the logs here and also reach out to Support on this? Thanks!
us-mdn-06-az05-e_diagnostics_20260127-1523.zip (904.8 KB)
I grabbed this diagnostic bundle from the gateway while I could (before it crashed again), thank you!