The Average_Roundtrip_Latency_ms in show leaves output is going to a 13 digit value in just 1 second gap. memsql version : 8.1.7
There is no latency observed between the servers , ie, master aggregator and leaf nodes.
The value used to come down to 1 ms in another 30 to 50 seconds.
There is no impact noticed in the cluster behavior and health due to this. This observed towards all the leaves at random timings as well.
Can we get an explanation on this unexpected hike.
Attaching the screen shot of show leaves command along with time at which query output is taken.
That looks like a bug. I’m going to file a bug with our engineering team on this. But probably they will need more details to reproduce it. Can you describe the details of your configuration and how you first observed this, and (if possible) how you can reproduce it? If you are a paying customer, you can open a support ticket if you like to allow the support team to help gather the repro.
If it does not clear itself, consider restarting your cluster or workspace. I’d also recommend upgrading to the latest patch of 8.1 or even upgrade to 8.5.
Thanks for reporting this. That looks like a bug for sure. Could you please check if querying the same information from information_schema.LEAVES table outputs a correct value? It can be used as a workaround if it does, and it would really help if you could share that output.
We have upgraded to 8.5.13 and the issue still persists.
Also we have deployed one more cluster with 8.5.13 and no issue is observed there.
As the value is quiet high it cannot be actual latency , but something wrongly calculated.
Also no other queries are facing any delay for execution.
The output of information_schema.LEAVES also returing same output. Please find the results attached.
