Chart connection lost to database intermittently

Chart becomes red & disconnects from database intermittenlty. It becomes normal after 4-5 seconds. What may be the reason. Is it java problem??? It happens in Linux as well as WIndows Clients. We have the Ignition 7.9 version running on Linux Ubuntu Server with Postgres database.

How many queries/sec are you running? I have seen charts disconnect from my database before and it basically came down to the database connection faulting because the gateway was trying to push too much data into the alarm events table (a fault on my setup by the way).

Check to see if you see any error messages on the gateway console.

Are you charting from the tag historian, or other data (transaction groups or other sources)? If not from the tag history system, it’s extremely important to create useful indices on the columns of your data that are used in the WHERE and ORDER BY clauses of the chart’s queries.

1 Like

Ofcourse I am charting from the tag historian…

Is there any limit on queries/sec???

In theory you are only limited by your hardware!
You probably need to tune your database installation, if you’ve just installed it and kept the default settings it’s performance will most likely be poor.
Check this site out for more details:

Thank you it is a very useful link…

There is not a hard set limit. The more you are running, the harder your gateway has to work. I think Inductive has told me they have ran over 100 queries/sec without any problems.

Ever since I realized the power of Transaction Groups I never looked back… I still don’t use tag historian :sunglasses: not even once ( is that bad? :stuck_out_tongue:)
Only time I’ve seen issues with chart connection was either there was an issue with my internet connection or there was a DB connection issue or if the query doesn’t return the result that works with the chart and therefore it freaks out and turns red out of shame of not being able to handle the data ( sarcasm lol).

ThanQ for reply… I will try with Transaction Group…

I feel like we still haven’t figured out whats going on with the chart acting up like that because if it is connection issue, using Transaction groups won’t fix the issue. Don’t get me wrong it will make your life so much easier but still need to figure out why the chart is losing connection.
does it happen often? or just once a while?

It happens frequently… But one thing I have observed data is not lost… I could able to get the data during that time interval…

This sort of problem is often an overloaded or poorly optimized database, where the slug of data for the chart sometimes exceeds the timeout in the client configuration. It is critically important to add useful indices to all tables, especially on t_stamp columns.