kentonv changed the topic of #sandstorm to: Welcome to #sandstorm: home of all things sandstorm.io. Say hi! | Have a question but no one is here? Try asking in the discussion group: https://groups.google.com/group/sandstorm-dev | Public logs at https://botbot.me/freenode/sandstorm/
afuentes has joined #sandstorm
TimMc has quit [Ping timeout: 240 seconds]
TimMc has joined #sandstorm
n8a has quit [Ping timeout: 252 seconds]
afuentes has quit [Ping timeout: 276 seconds]
n8a has joined #sandstorm
TimMc has quit [Ping timeout: 256 seconds]
TimMc has joined #sandstorm
TimMc has quit [Ping timeout: 256 seconds]
TimMc has joined #sandstorm
pie_ has quit [Ping timeout: 248 seconds]
patrickod_ has joined #sandstorm
M-hrjet has quit [*.net *.split]
JonTheNiceGuy has quit [*.net *.split]
eternaleye has quit [*.net *.split]
thejohnhenry[m] has quit [*.net *.split]
patrickod has quit [*.net *.split]
eldritch has quit [*.net *.split]
patrickod_ is now known as patrickod
<kentonv> so the Sandstorm release that just went out includes the HTTP proxy rewrite. It's currently off by default, but if you want to try it, add `EXPERIMENTAL_GATEWAY=true` to your sandstorm.conf and restart
<kentonv> the effect should be lower CPU and RAM usage... hopefully no other visible change.
ripdog has quit [Ping timeout: 240 seconds]
eldritch has joined #sandstorm
ripdog has joined #sandstorm
thejohnhenry[m] has joined #sandstorm
<kentonv> also if you use Sandcats TLS, with this setting, the SMTP port will now support STARTTLS -- however, sandcats doesn't publish MX records IIRC so this is probably moot (for now)
eternaleye has joined #sandstorm
M-hrjet has joined #sandstorm
JonTheNiceGuy has joined #sandstorm
TimMc has quit [Ping timeout: 256 seconds]
TimMc has joined #sandstorm
<kentonv> the new HTTP stack is now on for Oasis. This also means that nginx is no longer running on Oasis production servers.
harish has quit [Ping timeout: 256 seconds]
Lionel_Debroux_ has joined #sandstorm
Lionel_Debroux has quit [Ping timeout: 240 seconds]
n8a has quit [Ping timeout: 276 seconds]
n8a has joined #sandstorm
n8a has quit [Ping timeout: 252 seconds]
n8a has joined #sandstorm
n8a has quit [Ping timeout: 252 seconds]
TimMc has quit [Ping timeout: 256 seconds]
TimMc has joined #sandstorm
n8a has joined #sandstorm
afuentes has joined #sandstorm
Telesight has joined #sandstorm
harish has joined #sandstorm
harish has quit [Ping timeout: 265 seconds]
TimMc has quit [Ping timeout: 256 seconds]
TimMc has joined #sandstorm
pie_ has joined #sandstorm
dwrensha_ has joined #sandstorm
dwrensha has quit [Quit: Lost terminal]
TimMc has quit [Ping timeout: 256 seconds]
TimMc has joined #sandstorm
ccx^xmpp has left #sandstorm ["Disconnected: Received SIGTERM"]
ccx^xmpp has joined #sandstorm
afuentes has quit [Read error: Connection reset by peer]
TimMc has quit [Ping timeout: 240 seconds]
TimMc has joined #sandstorm
taktoa has joined #sandstorm
Lionel_Debroux_ has quit [Ping timeout: 240 seconds]
Lionel_Debroux has joined #sandstorm
Lionel_Debroux has quit [Ping timeout: 240 seconds]
Lionel_Debroux has joined #sandstorm
Lionel_Debroux has quit [Ping timeout: 240 seconds]
TimMc has quit [Ping timeout: 260 seconds]
TimMc has joined #sandstorm
n8a has quit [Ping timeout: 252 seconds]
n8a has joined #sandstorm
taktoa has quit [Remote host closed the connection]
Lionel_Debroux has joined #sandstorm
TimMc has quit [Ping timeout: 260 seconds]
TimMc has joined #sandstorm
TimMc has quit [Ping timeout: 256 seconds]
TimMc has joined #sandstorm
Telesight has quit [Remote host closed the connection]
isd has joined #sandstorm
taktoa has joined #sandstorm
<kentonv> dang... so far on Oasis the new gateway appears to be outperforming nginx alone, without even considering how much work it is offloading from Node.
TimMc has quit [Ping timeout: 256 seconds]
TimMc has joined #sandstorm
<JonTheNiceGuy> For stupid lay-people like me, is this because your new code is specifically written for oasis, or because nginx is bloated, or something else? Will this go live into sandstorm normal?
<isd> JonTheNiceGuy: (a) it's already in the main repo, but disabled for now because it's expiremental; you can enable it with a config option. (b), I think kentonv's remark is somewhat out of surprise; the email he sent out announcing it said he didn't expect it would do better than nginx
<kentonv> yeah, my code shouldn't be faster than nginx, it's not very optimized...
<isd> kentonv: does the gateway do websession to the app directly?
<isd> Also, This is just "server is serving even static pages faster", or...?
<kentonv> the gateway does WebSession directly, yes (currently only for UI sessions, not API, but adding API should be easy)
<kentonv> it forwards all traffic it doesn't explicitly recognize on to the shell (node.js process)
<kentonv> so pretty much it's serving the same purpose nginx used to: being an SSL-terminating reverse-proxy in front of the frontends. However, it additionally does WebSession proxying directly.
<kentonv> where "frontends" means "shell" -- I'm trying to teach myself to call it "shell" since "frontend" is clearly not the right word anymore