[12:26] kwmonroe: honestly, im not sure. I went through a rebuild thanks to some help with bdx, so im hoping its resolved. Ill update very shortly, im checking out the new build now [12:48] kwmonroe: ok so the update is Ive got rsyslog forwarding from....annnnd im an idiot [13:01] ok, so just to clarify, I send the logs to graylog, not to elasticsearch correct? [13:01] graylog is the logging server portion of graylog right? im not losing my mind here? [16:24] any sad souls hanging around on a weekend want haproxy charm options parsing fixed see here https://code.launchpad.net/~jamesbeedy/charm-haproxy/options_newline/+merge/345463 [16:55] do you ever stop working bdx? x) [16:57] TheAbsentOne: ahah, I have to catch up on bugs I hit during the week and didn't have time to tend to [16:57] either filing them or fixing or both [16:57] bdx much respect sir! [17:01] bdx: since you are here I can ask it, do you have experience with the pgbouncer charm? [17:02] ahh .. I haven't used it for a few years, but yeah kindof [17:03] looks like its actively maintained [17:03] and has docs and stuff [17:04] I heard the pgbouncer charm would allow me to set multiple authentication hosts in the charm itself, but I don't really see how [17:04] gonna have a look at the code [17:05] bdx: correct me if I'm wrong but shouldn't an interface allow this instead of a charm? [17:06] interfaces are really just for helping communicate information from one application to another .... [17:06] take it as you will [17:06] thats how I understand it [17:07] yeah exactly bdx [17:09] I fail to see how the bouncher charm would help me in adding an extre entry to the host auth file on the pgsql server [17:13] ahh [17:13] TheAbsentOne: Possibly because pgbouncer would be the one communicating with postgresql [17:13] so you just communicate with pgbouncer at that point [17:14] then postgresql only need a single entry in the pg_hba.conf for pgbouncer [17:14] I could be entirely wrong, but I seem to remember something to that degree [17:15] TheAbsentOne: https://gpdb.docs.pivotal.io/43180/admin_guide/access_db/topics/pgbouncer.html#pgb_config [17:16] ^ something similar to that may be going on under the covers, I'm really not familiar enough to say [17:17] but can I request that in my charm using pgbouncer bdx? [17:18] TheAbsentOne: I'm unsure what you are trying to do [17:21] yeah my bad I'll try to illustrate it: Assume I have a model with the postgresql charm and the pgbouncer charm deployed. Now assume that I have charm A that is a website that will access a postgres database and charm A is connect to charm B. The one connecting to (either postgresql or pgbouncer) is charm B however. This means that charm B is automatically configured in the host auth files but charm A's host is not allowed to conn [17:21] I want in charm B to write code that set's, "host A should be allowed to connect to the db as well" [17:21] does that makes sense bdx? [17:27] yeah [17:30] TheAbsentOne: what you are trying to do essentially is bypass part of the postgresql auth system [17:30] do you understand that? [17:31] well not really I want to pass a config string that the postgresql charm needs to handle; as a matter of fact there is an option for that in the postgresql but this means manual interaction which I want to evade [17:32] got it [17:32] yeah, thats how I've always gotten around it [17:32] the postgresql charm does "something" (<-- add an entry in the host auth file) for the requesting charm. I want the postgresql charm to do the same thing but for a host that I pass instead [17:32] setting extra_pg_auth [17:32] totally [17:32] yep that works ok I guess [17:33] but it would be awesome if there was a function that would allow it in the pgsql interface, or an optional parameter in the set_database method that requests a db and makes everything happen [17:33] "I want the postgresql charm to do the same thing but for a host that I pass instead" - got it, so just a custom entry to the pg_hba.conf [17:33] totally [17:33] I would make a feature request against the pgsql interface for that [17:33] I hope I can convince stub for something like that x) I can see it being usefull for others too [17:34] yeah, if you make a feature request/bug for it post it here and I'l give it some heat for you [17:34] I can see that being super handy [17:34] I talked about it though with stub it was he who said take a look at pgbouncer [17:35] The only problem is that I don't have a launchpad account and the pgsql interface is not hosted on github x) [17:35] TheAbsentOne: sounds like you know what you need to do then :) [17:35] or is there another way to request it formally? bdx as I haven't used launchpad before [17:36] yea, thats totally annoying, but people have the right to use what version control they please [17:37] totally bdx I'm creating an ubuntu one account as we speak [17:38] TheAbsentOne: that will be useful for many things, charmstore, JAAS, etc etc etc [17:38] but will also allow you to interact with the greater community too [17:40] Yeah thought so too but I hadn't the need for it so far. Could you point me where I can ask/feature request/issue on launchpad, it seems I'm blind [17:41] ahn wow was only look at the git.... [17:42] there should be a link on the charm webpage [17:42] to "submit a bug" [17:42] follow that and submit your feature request there [17:42] yeah found it, Ill just submit a bug idd, thanks for the help bdx [17:43] np [17:58] It is done, back to writing x) [18:18] TheAbsentOne: post it here [18:24] bdx: https://bugs.launchpad.net/interface-pgsql/+bug/1770885 [18:24] Bug #1770885: add support for multiple host auth when requesting db [18:26] TheAbsentOne: nice work [18:28] haha thanks, we'll see what happens and I hope it might help others too [18:44] kwmonroe: https://github.com/jamesbeedy/charm-graylog/blob/endpoints/reactive/graylog.py#L479,L522 - getting close [18:45] we will make this thing scale out and drop the apache2 bs here soon [18:58] nice code bdx I wish I had my eyes on things like that weeks ago! [18:59] I feel like I'm only startint to understand it right now and I feel stupid for it xD [18:59] TheAbsentOne: its cool man, everyone has their own learning curve [19:05] TheAbsentOne: fyi, I've been eyeing these changes for months [19:05] still don't have it perfect, but I'm real close [19:06] why I need a fe hours on the weekends to hack these types of things out [19:07] yeah it's very time consuming for some reason while it seems so little/easy at first [19:08] ^ programming [19:10] basicly xD but still, I underestimated reactive programming very hard [19:12] TheAbsentOne: its all relative :) [19:20] last bug [19:20] https://bugs.launchpad.net/charm-haproxy/+bug/1770890 [19:20] Bug #1770890: haproxy should wait to write out a front end and backend until they are configured by relating service [19:21] im off to get some sunshine now [19:21] lates [19:23] bdx enjoy!