Ant Wilson
Ant Wilson
Hey everyone thanks for your comments! To be clear this is _not_ an example of security through obscurity, the red herring here might be the presence of faux info or...
@naegelin I also agree with your idea that the admin endpoints should return truthful responses, if you are creating an account from the backend. We recently added [`auth.api.createUser`](https://github.com/supabase/gotrue-js/blob/0ae593b353c56487aaf0ea4084d5a3de89b37515/src/GoTrueApi.ts#L43) I'm not...
hey! Was just discussing this issue with @hf in depth - his opinion is that since it's still possible for people to determine the existence of accounts with side-channel attacks...
initial results: Some results using the get all: http.get(`${supabaseUrl}/rest/v1/read?select=*`, params) and write http.post(`${supabaseUrl}/rest/v1/write`, JSON.stringify(body), params) EC2 vCPU credit mode is Standard unless otherwise stated **t3a with 10 VUs over 30...
2xlarge Standard Mode I ran k6 across two processes against 2xlarge instances and the throughput remained the same overall (actually less than previously at 701/s reads)- just to confirm that...
### **Reads** 1 million rows/documents with random reads by equality on indexed column. 30s 10 VUs. Firestore `requests: 344/s http_req_duration: avg=28.6ms min=20.82ms med=26.48ms max=134.33ms` **c4.xlarge** Supabase direct to IP `requests:...
noticed that after inserting 1 million rows that supabase (beam) process was using ~100% CPU and ~50% memory for quite some minutes afterwards, it makes sense, just thought it was...
@steve-chavez have you benchmarked postgrest (nightly) recently. We're current on `https://github.com/PostgREST/postgrest/releases/download/nightly/postgrest-nightly-2020-11-10-22-47-3830887-linux-x64-static.tar.xz`
Think I've figured out the issue with reduced performance: **Cloudflare proxy performance has reduced dramatically** # Singapore to Singpore ### t3a.micro Unlimited via cloudflare proxy (supabase.co w/logflare worker) $ `top`...
@steve-chavez could you run these on a larger instance (maybe t3a.large) and see if you can get any improvement. For KPS I can't improve on ~1200/s ([have tried micro, medium,...