I keep getting problems accessing VSX over the last few days. The connections time out with “503 Service Unavailable” suggesting that the server is being hammered.
I tried connecting through different networks and different devices with the same result (just to check it is not me, although i can’t see how 503 could be) so it looks like either VSX is being used a lot recently or someone is either attacking it or not using it very efficiently…
Could someone please provide some details on what is going on?
It has been like this for a while and i’ve seen it happening on and off quite a few times over the last few days.
I just wonder if this is an attack on the servers or if someone has decided to download the whole database through the API one star at the time … in parallel
I note that VSX REST APIs are also impacted. I suppose this could be related to planned work. There’s no planned outage event in the event calendar though: AAVSO Programs & Events | aavso
After reading the forum posts here I spent most of my evening diagnosing the issue. I’ve tracked the issue down to a series of five queries that are overwhelming the database server. I’ve implemented a patch to the most expensive query that improves its performance by approximately 40x. I have a general idea where the other queries are located in code, but it will take me a while to track those down.
I’ve asked Aru to pull my code, test it, and deploy it when he wakes up tomorrow morning.