Page load time scales with peer/channel count on production nodes #113
ronballesteros
started this conversation in
General
Replies: 1 comment 11 replies
|
Hi @ronballesteros please test the branch or docker image v5.0.5 and let me know if it solved the problem |
11 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
Hi everyone...
We've been running peerswap-web on a production LND node with a decent number of peers and noticed the index page takes 30+ seconds to load after login, sometimes triggering a 504 from our load balancer. On a fresh node with no peers it loads near instantly.
Looking at the code,
indexHandlermakes several sequential gRPC calls withcontext.Background()(no timeout)...includingln.FeeReportwhich returns fee data for every channel on the node. Our hypothesis is the cumulative load time scales with peer/channel count, but we haven't added timing instrumentation to confirm which call is the actual bottleneck.Would love to know if this is a known issue or if there's a recommended approach before we dig deeper.
Environment:
Observed behavior:
On a production node with a significant number of peerswap peers and Lightning channels, the index page (and post-login redirect) takes 30+ seconds to load. On a fresh node with no peers the page loads near-instantly. Based on this, it took about 46 seconds...
However, Bbypassing any ingress and loadbalancer and directly connecting to the peerswap-ui pod still takes about 36 seconds to login.
On my dev/sandbox node with no peers, login happens instantly.
Each call takes longer as peer/channel count grows. With no peers these return instantly. With many peers the cumulative time easily exceeds 30 seconds, which also causes load balancer timeouts (504s) in hosted environments.
Thanks!
Ron
All reactions