*** jamesmcarthur has joined #openstack-keystone | 00:45 | |
*** jamesmcarthur has quit IRC | 01:13 | |
*** jamesmcarthur has joined #openstack-keystone | 01:20 | |
*** cmorpheus is now known as cmurphy | 01:24 | |
*** jamesmcarthur has quit IRC | 01:30 | |
openstackgerrit | Colleen Murphy proposed openstack/keystone master: [WIP] Add immutable resource option for roles https://review.opendev.org/666739 | 01:49 |
---|---|---|
openstackgerrit | Colleen Murphy proposed openstack/keystone master: Add --immutable-roles flag to bootstrap command https://review.opendev.org/675228 | 01:49 |
openstackgerrit | Colleen Murphy proposed openstack/keystone master: Add immutable roles status check https://review.opendev.org/675509 | 01:49 |
*** jamesmcarthur has joined #openstack-keystone | 02:12 | |
openstackgerrit | Colleen Murphy proposed openstack/keystone master: Add protection tests for trusts API https://review.opendev.org/675720 | 02:31 |
*** jamesmcarthur has quit IRC | 02:38 | |
*** jamesmcarthur has joined #openstack-keystone | 02:47 | |
*** mordred has quit IRC | 02:48 | |
*** mordred has joined #openstack-keystone | 02:48 | |
*** jamesmcarthur has quit IRC | 02:56 | |
*** jamesmcarthur has joined #openstack-keystone | 02:59 | |
*** jamesmcarthur has quit IRC | 03:25 | |
openstackgerrit | Colleen Murphy proposed openstack/keystone master: [WIP] Move hardcoded trust enforcement to default policies https://review.opendev.org/675807 | 04:15 |
cmurphy | kmalloc: beekneemech interested in your thoughts on ^ on monday | 04:15 |
openstackgerrit | Colleen Murphy proposed openstack/keystone master: [WIP] Add immutable resource option for roles https://review.opendev.org/666739 | 04:19 |
openstackgerrit | Colleen Murphy proposed openstack/keystone master: Add --immutable-roles flag to bootstrap command https://review.opendev.org/675228 | 04:19 |
openstackgerrit | Colleen Murphy proposed openstack/keystone master: Add immutable roles status check https://review.opendev.org/675509 | 04:19 |
openstackgerrit | Merged openstack/keystone master: Add API changes for app cred access rules https://review.opendev.org/628168 | 04:36 |
*** pcaruana has joined #openstack-keystone | 05:29 | |
*** pcaruana has quit IRC | 05:37 | |
*** jamesmcarthur has joined #openstack-keystone | 05:47 | |
*** pcaruana has joined #openstack-keystone | 05:50 | |
*** dancn has joined #openstack-keystone | 05:51 | |
*** jamesmcarthur has quit IRC | 06:03 | |
*** jamesmcarthur has joined #openstack-keystone | 06:06 | |
*** jamesmcarthur has quit IRC | 06:21 | |
*** markvoelker has joined #openstack-keystone | 06:31 | |
*** markvoelker has quit IRC | 06:36 | |
*** jamesmcarthur has joined #openstack-keystone | 06:45 | |
*** jamesmcarthur_ has joined #openstack-keystone | 06:51 | |
*** jamesmcarthur_ has quit IRC | 06:53 | |
*** jamesmcarthur_ has joined #openstack-keystone | 06:53 | |
*** jamesmcarthur_ has quit IRC | 06:53 | |
*** jamesmcarthur_ has joined #openstack-keystone | 06:54 | |
*** jamesmcarthur has quit IRC | 06:54 | |
*** jamesmcarthur_ has quit IRC | 07:04 | |
*** jamesmcarthur has joined #openstack-keystone | 07:05 | |
*** ivve has joined #openstack-keystone | 07:05 | |
*** jamesmcarthur has quit IRC | 07:10 | |
*** rcernin has quit IRC | 07:13 | |
*** tesseract has joined #openstack-keystone | 07:14 | |
openstackgerrit | Merged openstack/keystone master: Update api-ref for revocation list OS-PKI https://review.opendev.org/675296 | 07:53 |
*** tesseract has quit IRC | 08:42 | |
*** dancn has quit IRC | 08:42 | |
*** irclogbot_1 has quit IRC | 08:42 | |
*** hogepodge has quit IRC | 08:42 | |
*** mordred has quit IRC | 08:42 | |
*** dancn has joined #openstack-keystone | 08:42 | |
*** mordred has joined #openstack-keystone | 08:43 | |
*** tesseract has joined #openstack-keystone | 08:43 | |
*** openstackgerrit has quit IRC | 08:45 | |
*** irclogbot_1 has joined #openstack-keystone | 08:47 | |
*** irclogbot_1 has quit IRC | 08:49 | |
*** irclogbot_1 has joined #openstack-keystone | 08:52 | |
*** jaosorior has joined #openstack-keystone | 09:07 | |
*** mvkr has joined #openstack-keystone | 09:47 | |
*** xek has joined #openstack-keystone | 09:57 | |
*** markvoelker has joined #openstack-keystone | 10:09 | |
*** markvoelker has quit IRC | 10:14 | |
*** pcaruana has quit IRC | 10:43 | |
*** pcaruana has joined #openstack-keystone | 10:43 | |
*** jaosorior has quit IRC | 10:54 | |
*** whoami-rajat has joined #openstack-keystone | 11:08 | |
*** openstackgerrit has joined #openstack-keystone | 11:32 | |
openstackgerrit | Jens Harbott (frickler) proposed openstack/python-keystoneclient master: Stop using an admin endpoint by default https://review.opendev.org/675870 | 11:32 |
frickler | cmurphy: kmalloc: ^^ that's what it would take to make heat work in my devstack patch, but the tests blow up awfully and I'm unsure about backwards compatibility, pls let me know what you think | 11:34 |
frickler | see also https://review.opendev.org/675778 | 11:35 |
*** raildo has joined #openstack-keystone | 11:48 | |
*** jaosorior has joined #openstack-keystone | 12:03 | |
*** pas-ha has quit IRC | 12:13 | |
*** pas-ha has joined #openstack-keystone | 12:14 | |
*** aprice has quit IRC | 12:14 | |
*** aprice has joined #openstack-keystone | 12:14 | |
*** markvoelker has joined #openstack-keystone | 12:28 | |
*** jroll has quit IRC | 13:01 | |
*** jroll has joined #openstack-keystone | 13:02 | |
*** frickler has quit IRC | 13:18 | |
*** mchlumsky has joined #openstack-keystone | 13:25 | |
*** mchlumsky has quit IRC | 13:52 | |
*** mchlumsky has joined #openstack-keystone | 13:53 | |
*** beekneemech is now known as bnemec | 13:54 | |
*** atmark has joined #openstack-keystone | 14:00 | |
openstackgerrit | Colleen Murphy proposed openstack/keystone master: Add immutable roles status check https://review.opendev.org/675509 | 14:18 |
*** dancn has quit IRC | 14:26 | |
*** dancn has joined #openstack-keystone | 14:32 | |
openstackgerrit | Merged openstack/oslo.policy master: Move doc related modules to doc/requirements.txt https://review.opendev.org/669427 | 14:42 |
openstackgerrit | Radosław Piliszek proposed openstack/keystone master: Honor group_members_are_ids for user_enabled_emulation https://review.opendev.org/674782 | 14:43 |
yoctozepto | finally found time to dig into keystone's unit tests ^ | 14:43 |
*** petitbois has joined #openstack-keystone | 14:53 | |
petitbois | hey all -- not sure of the right place to ask this but i am having trouble authenticating my keystone object in python. I'd like to grab a user name from a user ID. Here's what I've got: http://paste.openstack.org/show/756341/ | 14:53 |
cmurphy | thanks yoctozepto | 15:11 |
cmurphy | petitbois: answered in -dev | 15:11 |
yoctozepto | cmurphy: yw; you had some doubts on the mailing list iirc - is everything clear now? | 15:13 |
cmurphy | yoctozepto: i still have some doubts, that ldap code is pretty dense and it's not clear to me how we didn't see this problem before, i'll probably be relying on people like raildo and gyee to help verify it | 15:16 |
yoctozepto | cmurphy: yeah, the code could use some cleanup, hence sprinkled TODOs | 15:19 |
yoctozepto | though I have yet to see ldap-facing code that I could describe as 'clean' | 15:20 |
yoctozepto | could be this old technology was not meant to ever be interfaced 'cleanly' :-) | 15:20 |
*** dave-mccowan has joined #openstack-keystone | 15:20 | |
raildo | cmurphy, ack, I'll review it util eod | 15:21 |
cmurphy | thanks raildo | 15:21 |
mnaser | so i'm having fun with ldap recently. | 15:22 |
yoctozepto | mnaser: "fun" you mean | 15:22 |
mnaser | right | 15:24 |
mnaser | for some reason keystone is not getting any records back and i'm wondering if its because its requesting users that have "userPassword" attribute | 15:24 |
mnaser | and this LDAP server does not supply it | 15:24 |
yoctozepto | mnaser: run debugging to see the queries, helped me | 15:26 |
mnaser | gah | 15:27 |
mnaser | of course | 15:27 |
mnaser | "dn: CN=XXXX,OU=Users,OU=$LOCATION,DC=$CORP,DC=com" | 15:27 |
mnaser | why would they not do CN=XXX,OU=$LOCATION,OU=USERS,DC=$CORP,DC=com | 15:28 |
* mnaser shrugs | 15:28 | |
mnaser | i wonder if that was it the whole time i'm ashamed :< | 15:28 |
yoctozepto | mnaser: change user root dn | 15:29 |
mnaser | yeah, i think that was it, i thought it was like some sort of tagged system | 15:29 |
mnaser | so i had it saying like | 15:29 |
yoctozepto | there is a lot of magic in this ldap code | 15:29 |
mnaser | OU=Users,DC=$CORP,DC=com | 15:29 |
mnaser | i think it might have done it | 15:29 |
yoctozepto | not to mention string interpolation domination | 15:29 |
yoctozepto | which makes it hard to read and search through | 15:29 |
mnaser | great now i dont have access to the server that it gets referred to | 15:31 |
mnaser | `ldap_chase_v3referrals` seems to run | 15:31 |
mnaser | https://bugs.launchpad.net/keystone/+bug/1233365 hrmmm | 15:32 |
openstack | Launchpad bug 1233365 in OpenStack Identity (keystone) "LDAP backend fails when connecting to Active Directory root DN" [High,Fix released] - Assigned to Dolph Mathews (dolph) | 15:32 |
yoctozepto | mnaser: almost 6 years since that | 15:36 |
mnaser | yoctozepto: doing an ldapsearch seems to give me all the users from ldap with same filter that keystone uses, but user list shows 4.. | 15:37 |
mnaser | result: 4 Size limit exceeded | 15:37 |
mnaser | oh hm | 15:37 |
mnaser | so i take the exact filterstr and attrs in ldapsearch on the same server | 15:45 |
mnaser | and it returns a result | 15:45 |
mnaser | but keystone ldap debug => "ld 0x38d2aa0 response count 0" | 15:45 |
mnaser | base=DC=<corp>,DC=com scope=1 filterstr=(&(&(sAMAccountName=<user>))(objectClass=organizationalPerson)(employeeID=*)) attrs=['employeeID', 'enabled', 'sAMAccountName', 'mail', 'userPassword', 'cn'] | 15:46 |
mnaser | ldapsearch -x -b "dc=<corp>,dc=com" -D "<snip>" -H ldaps://<snip>:636 -W "(&(&(sAMAccountName=<user>))(objectClass=organizationalPerson)(employeeID=*))" employeeID enabled sAMAccountName mail userPassword cn | 15:47 |
mnaser | the manual ldapsearch returns a result :< | 15:47 |
*** petitbois has quit IRC | 15:49 | |
yoctozepto | mnaser: scope=1 is single level | 15:51 |
yoctozepto | ldapsearch probably did scope=2 | 15:51 |
yoctozepto | for subtree | 15:52 |
yoctozepto | it is configurable in keystone | 15:52 |
*** dancn has quit IRC | 15:52 | |
mnaser | yoctozepto: yes indeed, darn ok let me try it out | 15:54 |
*** gyee has joined #openstack-keystone | 15:54 | |
*** dave-mccowan has quit IRC | 15:56 | |
*** dtruong has quit IRC | 15:59 | |
*** problem_v has joined #openstack-keystone | 16:00 | |
*** dtruong has joined #openstack-keystone | 16:00 | |
kmalloc | yoctozepto, mnaser: you can blame ayoung for most of the LDAP code ;). It's aweful to deal with (not adam's fault, but LDAP in python is gross...wait...strike "python" from that statement) | 16:34 |
*** dklyle_ is now known as dklyle | 16:34 | |
yoctozepto | kmalloc: yeah, it's gross anywhere you go, it lacks modern approach in libs | 16:36 |
kmalloc | yoctozepto: well pyldap is actually pretty good, but since we lean on python-ldap and ldappool...it's not so easy to convert | 16:38 |
kmalloc | or wait was it not pyldap, it was ldap3... that was the lib | 16:39 |
kmalloc | pyldap was the fork =/ | 16:39 |
kmalloc | ldap3 is pretty darn good | 16:39 |
yoctozepto | kmalloc: never tried anyway, my ldap journey was mostly with ldapjs in node.js and it was clumsy | 16:39 |
kmalloc | yoctozepto: i find a lot of <talk some protocol that isn't HTTP(ish)> in node/javascript is clumsy at best | 16:40 |
openstackgerrit | Colleen Murphy proposed openstack/keystone master: Move list_trusts enforcement to default policies https://review.opendev.org/675807 | 16:40 |
yoctozepto | kmalloc: nah, database access is pretty good (think sql/redis/mongodb/other-nosql-but-still-noldap-friends) | 16:40 |
*** ivve has quit IRC | 17:27 | |
*** whoami-rajat has quit IRC | 17:38 | |
*** jamesmcarthur has joined #openstack-keystone | 17:53 | |
*** atmark has quit IRC | 17:59 | |
*** atmark has joined #openstack-keystone | 18:00 | |
*** markvoelker has quit IRC | 18:13 | |
*** markvoelker has joined #openstack-keystone | 18:20 | |
*** dancn has joined #openstack-keystone | 18:24 | |
*** rafaelweingartne has joined #openstack-keystone | 18:27 | |
rafaelweingartne | Hello Guys, I am creating a Keystone python client using: "ks_client_v3.Client(session=session, trust_id=trust_id,....."Is it possible to use this client to generate tokens on the fly? Something similar to "openstack token issue" | 18:28 |
*** jamesmcarthur has quit IRC | 18:30 | |
openstackgerrit | Alex Schultz proposed openstack/keystoneauth master: Cleanup session on delete https://review.opendev.org/674139 | 18:31 |
*** jamesmcarthur has joined #openstack-keystone | 18:32 | |
*** jamesmcarthur has quit IRC | 18:34 | |
*** jamesmcarthur has joined #openstack-keystone | 18:34 | |
kmalloc | rafaelweingartne: are you trying to generate tokens that are extracted and used by a non-python application or something wrapped around keystoneclient? | 18:35 |
rafaelweingartne | I want to use by a python application | 18:36 |
kmalloc | rafaelweingartne: if you are looking for something wrapped around keystoneclient, I recommend using keystonauth instead (works like requests): https://docs.openstack.org/keystoneauth/latest/using-sessions.html | 18:36 |
rafaelweingartne | it is something I am creating for ceilometer | 18:36 |
kmalloc | rafaelweingartne: keystoneauth offloads a lot of the work for you and works just like requests and handles token creation/renewal/etc | 18:36 |
rafaelweingartne | I do not actually want a wrapper, I would like to execute the raw HTTP request with a token | 18:36 |
kmalloc | right. keystoneauth allows you to work just like requests does and handles openstack specific discovery/catalog parsing if you want | 18:38 |
kmalloc | keystoneauth is what openstack client and most other *-client libraries use. | 18:38 |
kmalloc | and what the SDK uses to implement access to openstack services | 18:38 |
rafaelweingartne | hmm | 18:38 |
rafaelweingartne | I am not sure I follow | 18:38 |
kmalloc | keystoneauth is a library that is based on python requests. | 18:38 |
rafaelweingartne | I mean, I will need to generate an HTTP request programmatically, and for that, I would need a token | 18:39 |
kmalloc | we bake into it's session code authentication (most forms), etc. | 18:39 |
*** markvoelker has quit IRC | 18:39 | |
rafaelweingartne | I do have the keystone client, but I am not sure how to use it to issue a token that I can use in the header of my HTTP request | 18:39 |
kmalloc | ok, lets step back. | 18:39 |
rafaelweingartne | Sure | 18:40 |
kmalloc | you're writing a python app that communicates with openstack services (ceilometer)? and what does it use to issue the requests to http? | 18:40 |
kmalloc | httplib? raw socket? curl (popen), python-requests? | 18:40 |
*** markvoelker has joined #openstack-keystone | 18:40 | |
rafaelweingartne | httplib | 18:41 |
kmalloc | ok. i highly recommend using python-requests if you can over straight httplib. | 18:41 |
kmalloc | it is, frankly, more feature complete | 18:41 |
kmalloc | and is generally super easy to use in comparison | 18:42 |
rafaelweingartne | is python requests the one you get when importing "requests"? | 18:42 |
kmalloc | yes | 18:42 |
rafaelweingartne | so, I am using the latter | 18:42 |
rafaelweingartne | sorry for the confusion | 18:42 |
kmalloc | no worries :) | 18:42 |
kmalloc | if you're using requests, keystoneauth implements all the auth workflow on top of requests and is a potential drop in replacement | 18:42 |
rafaelweingartne | hmm | 18:42 |
kmalloc | https://docs.openstack.org/keystoneauth/latest/using-sessions.html you create a session (which is really a requests object with extra openstack logic) | 18:43 |
kmalloc | and can either pass it to a current client (e.g. ceilometer-client) or you can make direct http calls via session.get() etc | 18:44 |
kmalloc | see https://docs.openstack.org/keystoneauth/latest/using-sessions.html#service-discovery | 18:44 |
kmalloc | as an example where you say "talk to identity and call /users" | 18:44 |
rafaelweingartne | ok, let me see if I understood | 18:44 |
kmalloc | it parses the catalog returned and allows working with the token (and handles adding the x-auth header) when communicating to the endpoint of your choice | 18:45 |
kmalloc | we created keystoneauth as a low-ish level library to offload the headache of auth in openstack, and we have plugins for most forms of auth (all new auth types end up with a plugin eventually) | 18:45 |
*** frickler has joined #openstack-keystone | 18:48 | |
rafaelweingartne | So, I just need to get the session and then issue the command "session_object.get("URL of the service I want to execute an HTTP GET request to")"? | 18:49 |
kmalloc | pretty much. it can either handle /parse from the token's catalog, or you can use it just like requests and pass a complete url. | 18:50 |
kmalloc | though if it's a known endpoint, i recommend using the discovery bits, it helps offload needing to know apriori two bits of data (the auth endpoint and the service endpoint) | 18:50 |
kmalloc | if you look at that link (the sessions one) it's pretty comprehensive on the forms you can use. | 18:51 |
kmalloc | and like i said, it's how openstack client, openstack SDK, and most other tools talk to openstack services. | 18:51 |
rafaelweingartne | that is cool | 18:51 |
rafaelweingartne | thanks!! | 18:51 |
kmalloc | heck yeah :) | 18:52 |
kmalloc | it's why we created the lib. saves lots of re-implemetation | 18:52 |
kmalloc | and it is something we heavily test | 18:52 |
kmalloc | also, note keystoneauth's API contract is extremely strict. | 18:52 |
kmalloc | barring a security reason, we will not break your use, even if it's unintended use of the lib | 18:52 |
kmalloc | basically, if we change out (even across major versions) behavior, we are wrong and i've reverted changes to ensure consistency | 18:53 |
rafaelweingartne | is the response a "response" object? | 18:53 |
rafaelweingartne | or is it just the response body | 18:53 |
kmalloc | yes, it should be a requests response objecty | 18:53 |
kmalloc | object* | 18:53 |
rafaelweingartne | thanks | 18:54 |
kmalloc | sure thing | 18:54 |
*** ivve has joined #openstack-keystone | 18:58 | |
rafaelweingartne | Dure, it works! | 19:02 |
rafaelweingartne | dude* | 19:02 |
rafaelweingartne | Awesome | 19:02 |
*** jamesmcarthur has quit IRC | 19:08 | |
*** rafaelweingartne has quit IRC | 19:09 | |
kmalloc | rafaelweingartne: glad it made your life easier! | 19:09 |
openstackgerrit | Colleen Murphy proposed openstack/keystone master: Move list_trusts enforcement to default policies https://review.opendev.org/675807 | 19:29 |
*** tesseract has quit IRC | 19:32 | |
*** dave-mccowan has joined #openstack-keystone | 19:54 | |
*** raildo has quit IRC | 20:52 | |
*** pcaruana has quit IRC | 20:56 | |
*** xek has quit IRC | 21:02 | |
*** markvoelker has quit IRC | 21:24 | |
cmurphy | kmalloc: could i get you to look at https://review.opendev.org/675807 before i keep going on the trust policies | 21:51 |
kmalloc | looking | 21:52 |
kmalloc | cmurphy: i think i see one bug | 22:03 |
*** markvoelker has joined #openstack-keystone | 22:03 | |
kmalloc | cmurphy: it mostly looks ok | 22:03 |
cmurphy | kmalloc: good catch | 22:05 |
*** lbragstad has joined #openstack-keystone | 22:09 | |
*** vishwanathj has quit IRC | 22:10 | |
*** mchlumsky has quit IRC | 22:14 | |
*** lbragstad has quit IRC | 22:17 | |
*** dancn has quit IRC | 22:25 | |
*** dancn has joined #openstack-keystone | 22:30 | |
*** ivve has quit IRC | 22:33 | |
*** markvoelker has quit IRC | 22:35 | |
*** hoonetorg has quit IRC | 22:48 | |
*** hoonetorg has joined #openstack-keystone | 22:50 | |
*** hoonetorg has quit IRC | 22:57 | |
*** vishwanathj has joined #openstack-keystone | 22:58 | |
openstackgerrit | Colleen Murphy proposed openstack/keystone master: Add protection tests for trusts API https://review.opendev.org/675720 | 22:58 |
openstackgerrit | Colleen Murphy proposed openstack/keystone master: Move list_trusts enforcement to default policies https://review.opendev.org/675807 | 22:58 |
*** dancn has quit IRC | 23:00 | |
*** markvoelker has joined #openstack-keystone | 23:00 | |
*** hoonetorg has joined #openstack-keystone | 23:01 | |
*** vishwanathj has quit IRC | 23:03 | |
*** markvoelker has quit IRC | 23:05 | |
*** rcernin has joined #openstack-keystone | 23:11 | |
*** dave-mccowan has quit IRC | 23:17 | |
*** markvoelker has joined #openstack-keystone | 23:18 | |
*** dave-mccowan has joined #openstack-keystone | 23:37 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!