I'm running a Kusama node with a actual network speed ranging from over 100MB/s to 500MB/s on both sides. However, I'm facing issues with slow block propagation time and low upload/download bandwidth speeds(under 100kB/s) on my node, even though I have a gigabyte network. What could be causing this problem, and how can I optimize my node's performance to improve these issues?
2
-
Where geographically is your node running? You can see there are geographic clusters of nodes: telemetry.polkadot.io/#map/…– Bruno ♦Commented Aug 17, 2023 at 15:22
-
Even if you have a lot of bandwidth, the nodes you're connected to may not which can cause your bandwidth utilization to be a fraction of the available bandwidth. And as Bruno already alluded to, your geographical location is also a factor.– Aaro AltonenCommented Aug 18, 2023 at 5:04
Add a comment
|
1 Answer
- Do you use Paritydb? Do you sync an archive node (state history)?
- Do you use a nvme SSD disk?
- Did you check the metrics of your disk during sync (avg read/write speed)?
- What is your CPU?
-
I used polkachu snapshots DB and I use 1TB SSD NVME, it seems working well and cpu is 8core 16thread. sometimes, upload/download bandwidth speed is over 1MB/s when its high and under 100kB/s when it`s low. very volatile....... what can cause this....– jay pCommented Aug 25, 2023 at 3:45
-
- Do you use Paritydb or Rocksdb? Do you sync an archive node (state history)? Or provide the full CLI command that you use to run the nodeю What are average read/write speeds (utilization) of your disk during sync? What is the model of your CPU?– kogelerCommented Aug 28, 2023 at 5:54
-