Gateway timeout while opening databases page

Now we have 3 active nodes:
Region1 Zone1 <-18ms-> Region1 Zone2 <-60ms-> Region2 Zone3
Lease preferense set to Zone1 due to main load.
While opening database page Admin UI all tables counting slow (1 table per second) and stopped with error after 30 seconds. Same problem with Data Distribution and Zone Configs, it’s not opening due to 30000ms promise error.

If we are modifing our installation, placing 2 zones to the same DC (2ms) for testing purpose - this pages opens very fast.

Is there any way to adjust gateway timeout to keep it working with high latency?

Hey @OlegSotnikov,

Can you provide a screenshot of http://#/reports/network

How many databases and tables are you running in your cluster? What are the sizes?

Thanks,
Matt


1 DB image

Hi Oleg,

This is a current WIP for us to improve this experience.

Currently, we are making two calls for each table to get the information.

Based on your network latencies, Since node 3 is far from node’s 1 and 2. The information is taking a long time to propagate through the network for leaseholders that are on node 3, which is most likely contributing to > 30 Seconds response time and ultimately the page timing out.

I believe the solution is for us to make our calls more efficient, and that will solve the gateway timeout issue for you.

As far as I am aware, there is no way to adjust the gateway timeout.