we’ve been seeing pretty regular hangs for a few months too. first the whole app would hang, but more recently just new queries would hang, while the app and existing (maybe cached) queries serve fine. similar symptoms as described here.
(we’re on 0.29.* and now 0.30, bare jar on a single m3.medium EC2 instance behind an application ELB. affected all browsers and all databases we have in metabase, both postgres and bigquery. we were seeing hangs as often as once or more per day, under moderate load.)
disabling http/2 seemed like a far-fetched fix, but i tried it, and we’ve now been up for five days with no hangs. fingers crossed, it may have done the trick. thanks @je2ryw @mmollick @a13n et al!