<gemlog>
Is it just me with the frequent, automatic refreshes on oasis? Darn irritating.
<gemlog>
Sorry, that's for tinyrss
jadewang has quit [Ping timeout: 265 seconds]
<gemlog>
It's just a few minutes. I barely have time to read an article. When I come back to the oasis-tinyrss tab, it has reloaded and I have to drill down again.
<gemlog>
Just mentioning. I'll look up the dev of tinyrss I guess.
gemlog has left #sandstorm ["Konversation terminated!"]
ocdtrekkie has quit [Remote host closed the connection]
aldeka has quit [Remote host closed the connection]
aldeka has joined #sandstorm
ocdtrekkie has joined #sandstorm
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 244 seconds]
rgrinberg has quit [Quit: WeeChat 1.5]
rgrinberg has joined #sandstorm
amyers has joined #sandstorm
amyers has quit [Remote host closed the connection]
amyers has joined #sandstorm
jadewang has joined #sandstorm
jemc has joined #sandstorm
jadewang has quit [Ping timeout: 244 seconds]
bodisiw has joined #sandstorm
asmyers has joined #sandstorm
amyers has quit [Ping timeout: 250 seconds]
jemc has quit [Quit: WeeChat 1.4]
jemc has joined #sandstorm
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 252 seconds]
bodisiw has quit [Quit: This computer has gone to sleep]
niekie has quit [Quit: No Ping reply in 180 seconds.]
niekie has joined #sandstorm
asmyers has quit [Ping timeout: 244 seconds]
bodisiw has joined #sandstorm
asmyers has joined #sandstorm
frigginglorious has joined #sandstorm
jadewang has joined #sandstorm
frigginglorious has quit [Ping timeout: 240 seconds]
jadewang has quit [Ping timeout: 252 seconds]
isd has joined #sandstorm
asmyers has quit [Ping timeout: 244 seconds]
<isd>
So, at the end of last week I had a little program that was connecting to the API via /tmp/sandstorm-api, and acquiring a wake lock. Now when I run it that socket isn't there. Afaik nothing changed, and I'm running out of ideas.
<dwrensha>
isd: are you no longer using sandstorm-http-bridge?
<isd>
It's still there.
<dwrensha>
sandstorm-http-bridge should set up that socket when it launches
<isd>
dwrensha: that's what I said.
<dwrensha>
does the problem happen for all grains? or just existing grains? or just new grains?
<isd>
All.
<isd>
At this point I don't have any state from last week though, so can't really test against the grains that had been working
<dwrensha>
hm... it'd be nice if we supported something like `spk dev --strace` that would run strace on the grain
<isd>
Yeah. I'm going to post everything somewhere I can send you links to.
bodisiw has quit [Quit: This computer has gone to sleep]
<isd>
Since go can actually cross compile decently, I've been building host side (just go build) so the only thing that actually happens in the vm is launching it.
<isd>
executable needs to be called irc-idler for it to work
<dwrensha>
your app needs to accept the connection from sandstorm-http-bridge, and then maybe wait for a tiny bit, and then /tmp/sandstorm-api should exist, I think
<isd>
Yeah, my reading of that is it won't show up until I start making HTTP connections, so if I start re-trying it's just going to deadlock
<isd>
I could have sworn I didn't change anything from when it was working, but that's *obviously* not ture.
asmyers has joined #sandstorm
frigginglorious has quit [Quit: frigginglorious]
<isd>
hm. So launched the api stuff in a separate goroutine and put a 10-second sleep in front of it. Seems to work, in that stuff actually comes back on http, and the exected stuff shows up in the console, but I thought there was supposed to be some sort of notification thing for the wake locks?
<dwrensha>
yeah, the user should see a notification in the top bar by the bell icon
<isd>
That's my memory from before, but I don't see one.
<isd>
Is there a way to inspect whether an app actually has a wake lock?
<dwrensha>
it's possible that you're hitting a bug
bodisiw has joined #sandstorm
<dwrensha>
you could do `sandstorm mongo` and examine the Notifications and the ApiTokens collections
<isd>
Maybe. If it *doesn't* have a lock, how long would realisticaly expect it to go without being killed, assuming no one has a browser open?
<isd>
Will try that
<dwrensha>
A grain with no active client will be spun down within 3-5 minutes, if I remember correctly.
frigginglorious has joined #sandstorm
<dwrensha>
it looks like you're the first serious use of wakelocks, so I wouldn't be surprised if you found some bugs
<dwrensha>
*user
<isd>
All the more reason to sanity check.
<isd>
So, apiTokens is empty, and there is no notifications table. I'm not super familiar with mongo.
<dwrensha>
sounds like the app doesn't have a wakelock, than
<dwrensha>
*then
<isd>
Yeah, it got shut down.
benkard has joined #sandstorm
benkard has quit [Max SendQ exceeded]
benkard has joined #sandstorm
isd has quit [Quit: Leaving.]
asmyers has quit [Ping timeout: 250 seconds]
amyers has joined #sandstorm
<asheesh>
cristipp: Hey! If you still want help, I'd love to email you. I'm asheesh@sandstorm.io
<asheesh>
isd: You too, though maybe I know how to reach you. : )
benkard has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
neynah has joined #sandstorm
bodisiw has quit [Quit: Leaving]
rgrinberg has quit [Ping timeout: 246 seconds]
vasile has quit [Ping timeout: 250 seconds]
Eclead has joined #sandstorm
isd has joined #sandstorm
Eclead has quit [Client Quit]
raoulzecat has joined #sandstorm
raoulzecat has quit [Ping timeout: 246 seconds]
<Lord>
hmmm piwik in sandstorm can't track https site because my sandstorm doesn't run on https
<Lord>
is there a free way to have a wildcard cert ?
<isd>
Lord: in general, not that I know of. This is part of why sandcats.io exists.
amyers has quit [Ping timeout: 252 seconds]
isd has quit [Quit: Leaving.]
frigginglorious has quit [Quit: frigginglorious]
isd has joined #sandstorm
cstrahan has quit [Ping timeout: 244 seconds]
cstrahan has joined #sandstorm
frigginglorious has joined #sandstorm
jemc has quit [Ping timeout: 265 seconds]
<isd>
So, trying to track down why my app isn't getting its wake lock. not super familiar with mongodb. dwrensha suggested I look in there to see if the notification was there as part of figuring out whether I was actually getting the lock. It wasn't, but what's more there wasn't even a notifications collection at all.
vasile_ has joined #sandstorm
<dwrensha>
my theory was that it would be created lazily if needed
<isd>
Is that normal? if it were a relational database I'd be like "ahah! the schema isn't even right!", but I don't know how mongo works and whether it needs to explicitly be there. apiTokens is there but empty, so I assume they're different things.
<dwrensha>
isd: you could try running jparyani's test app on that server
<isd>
good call, I'll give that a shot
<mrdomino>
hrm, ruby dev on sandstorm still seems really slow
<zarvox>
mrdomino: you're probably right, but do you mind sharing some numbers?
<mrdomino>
working on it. it's been 10+ minutes and i haven't seen webrick come up on openproject-ce, which is leading me to suspect that it might be something other than a performance issue.
<zarvox>
I know dwrensha's FUSE layer perf improvements landed some time ago, but the FUSE layer + Ruby's penchant for looking in not-useful folders probably still add a painful amount of overhead
<mrdomino>
i'm currently tyring to get the app running outside of vagrant so i can poke at its process structure more closely
<mrdomino>
ioctl(7, SNDCTL_TMR_TIMEBASE or SNDRV_TIMER_IOCTL_NEXT_DEVICE or TCGETS, 0x7fff877fafc0) = -1 EROFS (Read-on
<mrdomino>
ly file system)
<zarvox>
oh, TCGETS
<zarvox>
it's probably all the gem requirements
<mrdomino>
is it worth asking why it's trying to TCGETS on a file it opened?
<zarvox>
detecting if it's a terminal or a seekable file, I'd guess?
<mrdomino>
so it's not frozen, meaning rails s is taking somewhere in the ballpark of 5x to 100x as long to come up as it was a few weeks ago (not sure exactly)
decipherstatic has quit [Ping timeout: 250 seconds]
<zarvox>
strace also slows down the traced process by a significant amount
<mrdomino>
yeah, but we're talking 10 minutes prior to starting the strace
<zarvox>
but if it's still doing a bunch of gem loading, I'd attribute that to Rails potentially adopting an even-worse gem path strategy
<zarvox>
daaaaaaaaang
<zarvox>
that's rough
<zarvox>
dwrensha: ^ anything you'd suggest for the above?