These are chat archives for atomix/atomix

1st
Mar 2016
Richard Pijnenburg
@electrical
Mar 01 2016 06:11
Uhg. Hard to get up so early after 2 weeks of doing nothing lol
Jordan Halterman
@kuujo
Mar 01 2016 06:12
I feel ya
Richard Pijnenburg
@electrical
Mar 01 2016 06:16
First meeting is at 9 but want to make a good impression by getting in early
Jordan Halterman
@kuujo
Mar 01 2016 06:17
haha indeed
I started going to work early too
not a morning person but that’s part of my responsibility as a senior engineer I suppose
make sure people see me when they get there
Richard Pijnenburg
@electrical
Mar 01 2016 06:18
Hehe yeah
Sometimes at my old job I would get in really early just to get some work done before anyone would get. Work in peace for a bit.
Jordan Halterman
@kuujo
Mar 01 2016 06:19
I know I actually love that part of it
nobody shows up until 10
I get a couple hours to get shit done before my innevitable day of meetings
Richard Pijnenburg
@electrical
Mar 01 2016 06:20
Yeah indeed
Jordan Halterman
@kuujo
Mar 01 2016 06:20
just gotta get used to it
Richard Pijnenburg
@electrical
Mar 01 2016 06:21
Years ago I worked in shifts. But that was when I was younger lol
Not sure I could do that again
Starting at 7 am.
Richard Pijnenburg
@electrical
Mar 01 2016 07:47
For pr #138 is there a way to alert the user if he's got something Co
Argh
If he's got a same resource but different config
Jordan Halterman
@kuujo
Mar 01 2016 07:48
Only by checking the configuration
Richard Pijnenburg
@electrical
Mar 01 2016 07:48
Ah okay
Jordan Halterman
@kuujo
Mar 01 2016 07:48
Just prevents resources from being reconfigured. The first configuration is applied to all instances. That configuration is sent to all clients so they can see what the actual configuration is
Richard Pijnenburg
@electrical
Mar 01 2016 07:49
Just thinking of user experience :)
Ah okay. I see
Jordan Halterman
@kuujo
Mar 01 2016 07:50
Indeed. It's not ideal, but better than a resource's configuration changing out of nowhere
Gotta get to bed to go to work early :-P
Richard Pijnenburg
@electrical
Mar 01 2016 07:54
Hehe okay bud. Catch you tomorrow.
Btw I totally agree resources shouldn't change after its been created. Just would be nice to give an error somewhere.
Jordan Halterman
@kuujo
Mar 01 2016 08:03
That should be a configuration option when creating the clister
Cluster*
Actually, Copycat's error reporting just isn't refined enough to do that well yet
It basically only supports a few static errors right now
Have to fix that to be able to throw specific exceptions on the client indicating a configuration conflict
demonti
@demonti
Mar 01 2016 12:44
Hi, I am just trying to build copycat from master branch (mvn package), but get stuck in the tests ("Running io.atomix.copycat.test.ClusterTest"). What I am doing wrong?
Richard Pijnenburg
@electrical
Mar 01 2016 12:46
Hi, what java version are you using? I haven't ran the tests recently but can do on my test box to see if i see the same.
demonti
@demonti
Mar 01 2016 12:50
Java(TM) SE Runtime Environment (build 1.8.0_65-b17)
on Linux, 64 bit
Richard Pijnenburg
@electrical
Mar 01 2016 12:52
running the test suite now and see what i get
demonti
@demonti
Mar 01 2016 12:54
thanks
Richard Pijnenburg
@electrical
Mar 01 2016 13:58
sorry for the late reply. had lunch. build worked for me. can you retry?
demonti
@demonti
Mar 01 2016 14:17
cleared ~/.m2/repository, but still doesn't work. Which Java version are you using?
I don't see any kind of error messages. There is some repetition in the logs, though.
demonti
@demonti
Mar 01 2016 14:23
(I have to admit that I am a complete Maven novice – we use classical Netbeans/Ant scripts for building our software)
Jordan Halterman
@kuujo
Mar 01 2016 16:27
The Copycat tests hang? The repetition you see is likely heartbeats in the servers. @demonti you can build it and skip the tests just to get it built: mvn clean package -DskipTests but the tests hanging are curious. Wish I could reproduce them
@demonti the tests should all have timeouts such that if they're hanging they will be killed after 10 or 30 seconds. There's a lot of repetition in the test logs as those are heartbeat requests being sent between servers. But the tests do take a while to run and should finish. It sometimes takes 10 minutes or so. How long did they run for?
Richard Pijnenburg
@electrical
Mar 01 2016 16:39
@demonti
java version "1.8.0_66"
Java(TM) SE Runtime Environment (build 1.8.0_66-b17)
Java HotSpot(TM) 64-Bit Server VM (build 25.66-b17, mixed mode)
that's what i'm using
demonti
@demonti
Mar 01 2016 16:47
Cannot remember exactly, but I am sure to have it run once over several minutes. With the -D option, the build works. I will rerun the build and see whether it will terminate after 10 minutes.
demonti
@demonti
Mar 01 2016 17:06
Indeed, I tried it twice and roughly after 7 minutes, the test completed. Sorry for bothering you with this.
Jordan Halterman
@kuujo
Mar 01 2016 17:07
nice… no problem we actually need to shorten the tests for local development. Those long complex tests should really only be run on Travis