| === bodie__ is now known as bodie_ | ||
| thumper | marcoceppi: *bzzt* wrong | 01:07 |
|---|---|---|
| thumper | marcoceppi: juju run runs within a hook context | 01:07 |
| thumper | marcoceppi: however, it isn't as easy as it should be | 01:07 |
| thumper | marcoceppi: I managed to execute juju run to get a relation config value using the CLI | 01:08 |
| thumper | but it requires a bit of prior understanding that isn't entirely obvious | 01:09 |
| === zz_CyberJacob is now known as CyberJacob | ||
| === CyberJacob is now known as Guest59898 | ||
| === Guest59898 is now known as zz_Guest59898 | ||
| === mup_ is now known as mup | ||
| === MerryChristmas is now known as benonsotfware | ||
| === benonsotfware is now known as benonsoftware | ||
| === dooferlad_ is now known as dooferlad | ||
| pmatulis | with HA, how does the juju master get associated with the mongodb master? is it random (among the initial one + the ones which are about to be created) or is it always the initial single state server? | 13:48 |
| rick_h_ | pmatulis: I'd assume it'd have to be the first ones. Until the others are brought up and sync'd they're not able to be a master. | 13:54 |
| pmatulis | rick_h_: makes sense | 13:54 |
| rick_h_ | pmatulis: I guess it's not final/official. I think natefinch might know more/better if I'm incorrect. | 13:55 |
| natefinch | pmatulis, rick_h_: Mongodb chooses a master, and we make that Juju server the master | 13:57 |
| natefinch | pmatulis, rick_h_: if there's only one state server, obviously it's the master. When you add more, mongodb chooses a new master through its magical and mysterious algorithm | 13:58 |
| pmatulis | natefinch: ah, the juju master always resides on the same host as the mongodb master? | 13:59 |
| natefinch | pmatulis: yep, just easier that way | 13:59 |
| pmatulis | natefinch: perfect | 13:59 |
| bloodearnest | so, it seems destroy-environment in manual provider doesn't invoke destroy-service on the environments services. Is this expected? | 14:59 |
| bloodearnest | means I need to remember to manuall destroy-service's first, before tearing down the env | 15:00 |
| skylerberg | Hi guys. I am starting to write my charm and I see that pretty every charm (that I have looked at) calls execd_preinstall at the start of the install hook. It is not clear to me after looking at the source code for this function what it does. | 18:25 |
| marcoceppi | skylerberg: this is really for our IS team, they do a special bit that adds an execd directory to call additional files | 18:58 |
| marcoceppi | it's how they add code for certs and stuff | 18:59 |
| skylerberg | marcoceppi: Alright. Thanks Marco. | 19:01 |
| === zz_Guest59898 is now known as CyberJacob | ||
| === CyberJacob is now known as Guest75471 | ||
| === Guest75471 is now known as CyberJacob | ||
| mbruzek | Hello marcoceppi. I am still having trouble getting amulet to deploy with this added binary. I tried deleting the .bzr directory, but then I got juju-deployer errors, and I have tried your commit trick and which didn't work. | 20:27 |
| mbruzek | marcoceppi: Any other options? | 20:28 |
| marcoceppi | mbruzek: wait until next week and try again? | 20:28 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!