Our gateway 7.6.4 has ran steadily for months until this week we noticed huge chunks of no data in the middleof the night on our easy charts. The gateway was ‘off’ apparently for a couple of hours before recovering with no human interaction. This has just happened again whilst we were working on it. The GUI showed the gateway and web server alternating between running and stopped. A reboot of the server appears to have got the gateway running again. I managed to take a copy of the wrapper and have attached it.
We are on Windows Server 2003, Jave 1.6.0_24
We do not have the mobile module.
Can anyone help with this as when this occurs we are losing production.
Wrapper.zip (486 KB)
Your log has numerous ClockDriftDetector events, with large (many seconds) halts. I didn’t look at all of them, but I saw many > 5 seconds, and several > 20 seconds.
This is almost certainly due to java garbage collection. The rest of the errors seem to be side effects of java pause-the-world events – broken OPC connectons. Since you are still on java 6, you have fewer options, but at the very least you should turn on GC logging so you get the details.
Look at this post for the settings needed in ignition.conf. You can’t use the G1GC, though. You need at least Java 7 for that.