| *** Piet has joined #openstack-lbaas | 00:07 | |
| *** minwang2 has quit IRC | 00:14 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Bypass listener check on cert deploy during create https://review.openstack.org/237707 | 00:18 |
|---|---|---|
| *** bana_k has quit IRC | 00:18 | |
| *** diogogmt has quit IRC | 00:21 | |
| *** yamamoto has joined #openstack-lbaas | 00:23 | |
| *** yamamoto has quit IRC | 00:42 | |
| *** bana_k has joined #openstack-lbaas | 00:52 | |
| *** kbyrne has quit IRC | 01:14 | |
| *** kbyrne has joined #openstack-lbaas | 01:15 | |
| *** diogogmt has joined #openstack-lbaas | 01:42 | |
| *** yamamoto has joined #openstack-lbaas | 01:42 | |
| *** yamamoto has quit IRC | 01:48 | |
| *** bana_k has quit IRC | 02:03 | |
| *** vivek-ebay has quit IRC | 02:06 | |
| *** minwang2 has joined #openstack-lbaas | 02:24 | |
| *** yamamoto has joined #openstack-lbaas | 02:30 | |
| *** KunalGan_ has joined #openstack-lbaas | 02:54 | |
| *** KunalGandhi has quit IRC | 02:58 | |
| openstackgerrit | Merged openstack/octavia: REST Amp Agent: Compare lower case MAC Addresses https://review.openstack.org/237418 | 03:00 |
| *** KunalGan_ has quit IRC | 03:35 | |
| *** numans has joined #openstack-lbaas | 04:06 | |
| *** numans has quit IRC | 04:19 | |
| *** pckizer has quit IRC | 04:31 | |
| *** pckizer has joined #openstack-lbaas | 04:33 | |
| *** diogogmt has quit IRC | 04:45 | |
| openstackgerrit | Brandon Logan proposed openstack/octavia: REST Amp Agent: Handle interfaces file too https://review.openstack.org/237421 | 05:00 |
| *** minwang2 has quit IRC | 05:03 | |
| *** blogan_ has joined #openstack-lbaas | 05:03 | |
| *** openstackgerrit has quit IRC | 05:16 | |
| *** openstackgerrit has joined #openstack-lbaas | 05:17 | |
| *** bana_k has joined #openstack-lbaas | 05:22 | |
| *** bana_k has quit IRC | 05:35 | |
| *** blogan_ has quit IRC | 05:36 | |
| *** blogan_ has joined #openstack-lbaas | 05:37 | |
| *** jschwarz has joined #openstack-lbaas | 05:46 | |
| *** blogan__ has joined #openstack-lbaas | 05:59 | |
| *** blogan_ has quit IRC | 06:01 | |
| *** numans has joined #openstack-lbaas | 06:03 | |
| *** evgenyf has joined #openstack-lbaas | 06:13 | |
| *** blogan_ has joined #openstack-lbaas | 06:29 | |
| *** nmagnezi has joined #openstack-lbaas | 06:32 | |
| *** blogan__ has quit IRC | 06:32 | |
| *** armax has quit IRC | 06:32 | |
| *** jschwarz has quit IRC | 06:51 | |
| *** kobis has joined #openstack-lbaas | 06:57 | |
| *** bana_k has joined #openstack-lbaas | 06:59 | |
| *** bana_k has quit IRC | 07:24 | |
| *** jongchoi has joined #openstack-lbaas | 07:30 | |
| *** blogan__ has joined #openstack-lbaas | 07:38 | |
| *** blogan_ has quit IRC | 07:41 | |
| openstackgerrit | Stephen Balukoff proposed openstack/octavia: Adding introduction and quick start docs https://review.openstack.org/232173 | 08:00 |
| *** numans has quit IRC | 08:41 | |
| *** jongchoi has quit IRC | 08:48 | |
| *** openstackgerrit has quit IRC | 09:01 | |
| *** amotoki has joined #openstack-lbaas | 09:01 | |
| *** openstackgerrit has joined #openstack-lbaas | 09:01 | |
| *** numans has joined #openstack-lbaas | 09:17 | |
| *** openstackgerrit has quit IRC | 09:31 | |
| *** openstackgerrit has joined #openstack-lbaas | 09:32 | |
| *** pckizer has quit IRC | 09:37 | |
| *** pckizer has joined #openstack-lbaas | 09:37 | |
| openstackgerrit | Bogdan Tabor proposed openstack/neutron-lbaas: Switch to using neutron.common.utils:replace_file() https://review.openstack.org/232454 | 09:46 |
| *** Brian_shang has quit IRC | 10:51 | |
| *** Brian_shang has joined #openstack-lbaas | 10:52 | |
| *** doug-fish has joined #openstack-lbaas | 11:20 | |
| *** rtheis has joined #openstack-lbaas | 11:25 | |
| *** ducttape_ has joined #openstack-lbaas | 12:09 | |
| *** openstackgerrit has quit IRC | 12:16 | |
| *** openstackgerrit has joined #openstack-lbaas | 12:17 | |
| *** yamamoto has quit IRC | 12:21 | |
| *** ducttape_ has quit IRC | 12:34 | |
| *** yamamoto has joined #openstack-lbaas | 13:14 | |
| *** mestery has quit IRC | 13:16 | |
| *** ducttape_ has joined #openstack-lbaas | 13:16 | |
| *** ducttape_ has quit IRC | 13:18 | |
| *** kiran-r has joined #openstack-lbaas | 13:21 | |
| *** mestery has joined #openstack-lbaas | 13:29 | |
| *** kiran-r has quit IRC | 13:30 | |
| *** ducttape_ has joined #openstack-lbaas | 13:50 | |
| *** jongchoi has joined #openstack-lbaas | 14:08 | |
| *** orion__ has joined #openstack-lbaas | 14:10 | |
| *** nmagnezi has quit IRC | 14:20 | |
| *** jongchoi has quit IRC | 14:28 | |
| *** minwang2 has joined #openstack-lbaas | 14:37 | |
| *** yamamoto has quit IRC | 14:46 | |
| *** Piet has quit IRC | 14:51 | |
| *** alejandrito has joined #openstack-lbaas | 15:01 | |
| *** orion___ has joined #openstack-lbaas | 15:01 | |
| *** alejandrito has quit IRC | 15:02 | |
| *** alejandrito has joined #openstack-lbaas | 15:02 | |
| *** f13o has joined #openstack-lbaas | 15:03 | |
| *** orion__ has quit IRC | 15:04 | |
| *** orion___ has quit IRC | 15:04 | |
| *** yamamoto has joined #openstack-lbaas | 15:09 | |
| *** diogogmt has joined #openstack-lbaas | 15:12 | |
| *** Kiall has quit IRC | 15:19 | |
| *** Kiall has joined #openstack-lbaas | 15:20 | |
| *** yamamoto_ has joined #openstack-lbaas | 15:29 | |
| *** armax has joined #openstack-lbaas | 15:29 | |
| *** yamamoto has quit IRC | 15:32 | |
| *** evgenyf has quit IRC | 15:46 | |
| dougwig | Hello from the airport. | 16:00 |
| johnsom | Leaving or Arriving? | 16:00 |
| xgerman | dougwig lives at the airport - welcome home | 16:01 |
| *** manjeets has joined #openstack-lbaas | 16:01 | |
| *** manishg has joined #openstack-lbaas | 16:03 | |
| xgerman | dougwig, blogan: https://wiki.openstack.org/wiki/Octavia/Weekly_Meeting_Agenda#Meeting_2015-10-21 | 16:03 |
| manishg | hi folks, I'm looking at LBaaS APIv2.0 and had a basic question. why does adding a member requires to have a port (https://wiki.openstack.org/wiki/Neutron/LBaaS/API_2.0#Add_a_New_Member_to_a_Pool). Isn't port already part of listener attributes which is associated with a pool (and pool_id is already passed in member-create). Or is it that pool could be associated with multiple members and yet when a real is | 16:04 |
| manishg | attached it could be connected to only one of them (and thus necessitating the need to pass port#)? | 16:04 |
| xgerman | I think port is optional and only need if your member is on a different subnet | 16:04 |
| blogan | manishg: the listener port is the frontend, the port in which traffic coming to be load balanced, whereas the member port is the ports in which teh loadbalancer will send the load balanced traffic to those members | 16:04 |
| blogan | xgerman: no its not, protocol_port is required | 16:05 |
| xgerman | ah, must have confused it with the port on pool | 16:05 |
| blogan | xgerman: no port on pool :) | 16:05 |
| xgerman | wasn’t there some proposal? | 16:05 |
| johnsom | port as in TCP port | 16:05 |
| xgerman | ah | 16:05 |
| blogan | yeah | 16:06 |
| xgerman | manishg terminology | 16:06 |
| manishg | blogan: doesn't listener have the protocol_port also ? https://wiki.openstack.org/wiki/Neutron/LBaaS/API_2.0#Create_a_Listener | 16:06 |
| *** numans has quit IRC | 16:07 | |
| blogan | manishg: yes but thats the port in which the load balancer is listening, with the vip | 16:07 |
| xgerman | 1) this is not the official API doc | 16:07 |
| manishg | so, looks like this is for the frontend (VIP port), and the real could use different one. | 16:07 |
| blogan | manishg: but the members themselves needt o also be listening to receive load balanced traffic | 16:07 |
| xgerman | http://developer.openstack.org/api-ref-networking-v2-ext.html | 16:08 |
| openstackgerrit | Phillip Toohill proposed openstack/octavia: Adding new network driver for containers https://review.openstack.org/197858 | 16:08 |
| manishg | blogan: got it. thanks! So this gives flexibility to consolidate frontend ports into single port in members or map to different ports. makes sense. thanks. | 16:08 |
| blogan | manishg: yes | 16:08 |
| openstackgerrit | Phillip Toohill proposed openstack/octavia: Updates for containers functionality https://review.openstack.org/199954 | 16:09 |
| *** Piet has joined #openstack-lbaas | 16:11 | |
| dougwig | xgerman: flight for Tokyo takes off five minutes before meeting. If you can, please have folks review the Lbaas summit ether pad. | 16:12 |
| manishg | blogan: another question related to port# , is it expected that all members of pool will use same set of protocol_ports? At least, that's how we deploy them, so curious if there are requirements to support otherwise (the current approach seems to accommodate heterogenous members [listening on various ports] also). With this approach a member identifier = REAL+port (so same REAL listening on multiple ports = | 16:13 |
| manishg | multiple members). | 16:13 |
| xgerman | it’s upon there | 16:13 |
| johnsom | dougwig Safe travels | 16:13 |
| manishg | xgerman: thanks for the link to the api doc. | 16:13 |
| xgerman | yeah, make sure to refer to that | 16:13 |
| dougwig | johnsom: ty | 16:13 |
| xgerman | mannish you can mix & match on the backend | 16:13 |
| xgerman | manishg | 16:14 |
| blogan | manishg: if you set up a server to listen on port 8080 and another on port 80, and you add thsoe 2 as members to a pool with those same ports, it'll still work | 16:14 |
| blogan | dougwig: are you flying out of san jose? | 16:14 |
| manishg | blogan: yep. If a VIP is setup with two frontend ports 8080 and 80, and I add a server (with IP, IP-1) to listen on 5050 and 50 then it would be two different members in table (=> needs two member-create), right? IP-1 / 5050 and IP-1 / 50 ? | 16:16 |
| xgerman | if your vip is listening on two port you need two listeners... | 16:17 |
| dougwig | blogan: Seattle | 16:18 |
| manishg | xgerman: ah, was looking at the other doc... that has a concept of a listener. | 16:19 |
| manishg | in the doc you sent, the protocol/port_number is part of vip itself. | 16:20 |
| xgerman | yep | 16:20 |
| manishg | so is the other doc https://wiki.openstack.org/wiki/Neutron/LBaaS/API_2.0 older (and hence deprecated) or something that is in the works? | 16:20 |
| xgerman | nope, the docs I sent are official and being maintained… the wiki not so much | 16:21 |
| xgerman | we have two APIs LBaaS V1 and V2… both are on the official docs (with V1 being deprecated for Liberty) | 16:21 |
| xgerman | so in V2: Loadbalancer gets the VIP; Listener the port; and then pool and then meber | 16:22 |
| blogan | dougwig: too bad, if you're flying ANA you might have been able to get on the R2D2 plane | 16:22 |
| xgerman | I am on JAL | 16:22 |
| blogan | i'm on ANA flying out of Houston so it'll take me like 4 hours to even get out of the US | 16:23 |
| manishg | xgerman: so with v2 API, looks like you could have a pool with two listeners (two ports) and a member could still use a single port (that is like in v1 two ports => two members isn't true). | 16:23 |
| xgerman | yep, with v2 you can do that ;-) | 16:24 |
| manishg | ok, got it. so with v2 also member really is IP + port. so if a server (REAL) wants to listen on two ports it is considered as two members (and so two member-creates with individual ports). | 16:25 |
| *** logan2 has quit IRC | 16:27 | |
| xgerman | well, you can have the members listen on one port and then just make the loadbalancer listen on two; or add the member twice with different ports... | 16:28 |
| manishg | xgerman: is lbv2 integrated with heat, do you know? current code in heat looks like it's using lbv1 and also it is assuming the creates are sync (I think). There was some work on async being done by blogan I think. Does the backend now support async/ sync or is it assumed to be sync. Some pointers are greatly appreciated. | 16:29 |
| blogan | manishg: thats in progress | 16:29 |
| xgerman | +1 | 16:29 |
| xgerman | bana_k is working on heat | 16:29 |
| blogan | manishg: the heat integration | 16:29 |
| blogan | manishg: what do you mean by async vs sync? | 16:29 |
| blogan | manishg: the drivers? | 16:30 |
| manishg | blogan: async / sync - the drivers . so after create API call returns is it assumed the backend is programmed by the driver(s)? or does it return a "create_pending" and the caller checks state before next operation. (e.g. if there was a need to do an update, etc.) | 16:31 |
| blogan | manishg: thats still driver dependent | 16:32 |
| blogan | manishg: i would like to just make it so that all drivers are async, but just have had higher priorities | 16:32 |
| manishg | blogan: ok. but from lbaas API point of view, what are the expectations? Does it support 'status == create_pending' etc, ? If async is supported then does the driver go back and update the db to reflect 'active' state (once backend is all set)? | 16:33 |
| blogan | manishg: yes it does for all drivers | 16:34 |
| blogan | manishg: its just that with some drivers the create request may immediately respond with a status of ACTIVE | 16:34 |
| blogan | but take longer | 16:34 |
| blogan | if that makes sense | 16:34 |
| blogan | most are async though so you'll get a PENDING_CREATE | 16:34 |
| blogan | manishg: for an async driver, it is the driver's responsibility to call a method "successful_completion" whenever the operation is finished successfully | 16:35 |
| blogan | which does the status updates | 16:35 |
| manishg | blogan: ah, okay. thanks! that helps. so driver could 1) just say 'ACTIVE' but not really program the backend yet, 2) say 'ACTIVE' and be sync (that is will block API and program backend 3) return PENDING and later change to ACTIVE when done programming. | 16:36 |
| blogan | what i want some day is to not have the drivers have to worry about that, and thats handled automatically | 16:36 |
| blogan | manishg: correct | 16:36 |
| blogan | manishg: ml2 does similar doesn't it? | 16:36 |
| blogan | manishg: meaning it automatically updates teh db, doesn't leave it to the drivers | 16:37 |
| manishg | blogan: cool. I'll look at the code. I think the 'successful_completion' idea is good. ml2 just supports sync right now due to some other expectations from nova side (for some of the operations). | 16:37 |
| manishg | blogan: although ml2 has been wanting to move to async model, there are other priorities :( | 16:38 |
| manishg | and it hasn't happened yet. we had a 2 day meetup where we discussed in detail what could be done, etc. we'll see how it goes. | 16:38 |
| blogan | manishg: ah okay, yeah i didnt want to the drievrs to have access to teh db as that would cause an inconsistent experience, but didn't have the time to do it the correct way sooo that was a safe middle ground | 16:38 |
| blogan | manishg: ha yeah i know how that goes | 16:38 |
| manishg | blogan: yeah, this is a safe option and allows the option to go async for the drivers. I'll see if ml2 can do this. | 16:39 |
| *** doug-fish has quit IRC | 16:40 | |
| *** kobis has quit IRC | 16:40 | |
| blogan | manishg: sounds good, let me know if you need anything else | 16:40 |
| *** doug-fish has joined #openstack-lbaas | 16:41 | |
| manishg | blogan: thanks a lot. this was very helpful. | 16:41 |
| *** KunalGandhi has joined #openstack-lbaas | 16:42 | |
| blogan | manishg: you were the guy who gave me my first run down on taskflow in atlanta i think | 16:42 |
| blogan | right? | 16:42 |
| blogan | not sure if you remember | 16:42 |
| *** amotoki has quit IRC | 16:42 | |
| manishg | blogan: yeah. and we hosted the ml2 meetup for async couple of weeks ago for ml2 here at yahoo. | 16:45 |
| *** doug-fish has quit IRC | 16:45 | |
| blogan | manishg: ah, i need to get more involved with ml2 honestly, its something we need to look into possibly using | 16:45 |
| manishg | blogan: ml2 is also thinking of using taskflow but it's going slow... | 16:46 |
| *** logan2 has joined #openstack-lbaas | 16:47 | |
| blogan | manishg: i will say I have come to the conclusion that i do not like taskflow, my opinion is not shared by everyone though | 16:51 |
| blogan | manishg: i like the functionality of it, but the readability, and ability to write the code easily is just too much of a pain | 16:52 |
| manishg | blogan: for ml2 I'm not sure what the option will be. I think there is a need to something async but not clear on what that would be. We did talk about TF quite a bit. but no conclusion. | 16:53 |
| manishg | blogan: yeah, some of the other folks in ml2 also expressed that there should another layer on top of TF that makes some of these tasks simpler. | 16:54 |
| *** doug-fish has joined #openstack-lbaas | 16:56 | |
| blogan | manishg: yeah, i'd rather just write readable, reusable code and write a decent rollback mechanism | 16:57 |
| blogan | taskflow is great at that, but i just have come to the opinion that its not worth the trouble for that | 16:57 |
| *** zigo_ has quit IRC | 16:59 | |
| manishg | blogan: we'll see how it goes in ml2 and what we decide. will keep you posted. | 16:59 |
| *** zigo has joined #openstack-lbaas | 17:00 | |
| *** TrevorV has joined #openstack-lbaas | 17:03 | |
| *** yamamoto_ has quit IRC | 17:17 | |
| *** bana_k has joined #openstack-lbaas | 17:19 | |
| *** yamamoto has joined #openstack-lbaas | 17:25 | |
| *** kfox1111 has quit IRC | 17:25 | |
| *** kfox1111 has joined #openstack-lbaas | 17:26 | |
| *** ducttape_ has quit IRC | 17:35 | |
| *** amotoki has joined #openstack-lbaas | 17:43 | |
| *** amotoki has quit IRC | 17:48 | |
| *** mestery has quit IRC | 17:52 | |
| *** yamamoto has quit IRC | 17:55 | |
| *** klindgren_ is now known as klindgren | 17:57 | |
| *** harlowja has quit IRC | 17:57 | |
| *** harlowja has joined #openstack-lbaas | 18:02 | |
| *** manishg has quit IRC | 18:05 | |
| *** ducttape_ has joined #openstack-lbaas | 18:11 | |
| rtheis | When deploying DevStack with q-lbaasv2 and Octavia enabled, is the q-lbaasv2 screen service necessary? Creating a load balancer seems to work fine without it running. | 18:16 |
| *** Piet has quit IRC | 18:17 | |
| *** mestery has joined #openstack-lbaas | 18:20 | |
| *** vivek-ebay has joined #openstack-lbaas | 18:22 | |
| *** vivek-ebay has quit IRC | 18:22 | |
| *** vivek-ebay has joined #openstack-lbaas | 18:27 | |
| *** Aish has joined #openstack-lbaas | 18:33 | |
| *** bharathm has joined #openstack-lbaas | 18:33 | |
| *** fnaval has joined #openstack-lbaas | 18:43 | |
| *** amotoki has joined #openstack-lbaas | 18:44 | |
| *** amotoki has quit IRC | 18:49 | |
| *** minwang2 has quit IRC | 18:50 | |
| *** yamamoto has joined #openstack-lbaas | 18:55 | |
| *** yamamoto has quit IRC | 19:00 | |
| blogan | rtheis: the qlbaasv2 screen is just for the agent, which is indeed not used when using Octavia | 19:02 |
| rtheis | blogan: thank you | 19:04 |
| rm_work | redrobot: please review my comments https://review.openstack.org/#/c/237807/1/specs/mitaka/neutron-lbaas-store-tls-passphrase.rst,cm | 19:20 |
| *** bharathm has quit IRC | 19:21 | |
| *** Aish has left #openstack-lbaas | 19:31 | |
| openstackgerrit | Brandon Logan proposed openstack/octavia: REST Amp Agent: Handle interfaces file too https://review.openstack.org/237421 | 19:37 |
| *** minwang2 has joined #openstack-lbaas | 19:40 | |
| *** Piet has joined #openstack-lbaas | 19:43 | |
| *** TrevorV has quit IRC | 19:45 | |
| *** harlowja has quit IRC | 19:50 | |
| *** crc32 has joined #openstack-lbaas | 19:58 | |
| *** bharathm has joined #openstack-lbaas | 20:02 | |
| *** Brian_shang has quit IRC | 20:08 | |
| *** Brian_shang has joined #openstack-lbaas | 20:08 | |
| *** fnaval_ has joined #openstack-lbaas | 20:10 | |
| *** fnaval has quit IRC | 20:12 | |
| *** crc32 has quit IRC | 20:16 | |
| *** crc32 has joined #openstack-lbaas | 20:22 | |
| *** crc32 has quit IRC | 20:26 | |
| *** crc32 has joined #openstack-lbaas | 20:27 | |
| *** manishg has joined #openstack-lbaas | 20:38 | |
| *** harlowja has joined #openstack-lbaas | 20:39 | |
| *** harlowja_ has joined #openstack-lbaas | 20:40 | |
| manishg | in v2.0 APIs member-create takes a subnet-id. If vip is already setup, pool is already setup then what is the purpose of the optional subnet-id? (there is no need for allocation of any new VIP, etc.) | 20:41 |
| rm_work | internal customer networks need to be plumbed | 20:42 |
| rm_work | ^^ generic answer | 20:42 |
| rm_work | I think there is a more detailed answer | 20:42 |
| rm_work | it's for the backend member | 20:42 |
| *** alejandrito has quit IRC | 20:42 | |
| rm_work | if you add one on a new subnet, it would not be reachable | 20:42 |
| rm_work | unless you specify that | 20:42 |
| openstackgerrit | Merged openstack/octavia: Bypass listener check on cert deploy during create https://review.openstack.org/237707 | 20:43 |
| *** harlowja has quit IRC | 20:44 | |
| *** amotoki has joined #openstack-lbaas | 20:45 | |
| manishg | rm_work: if I create VM on a neturon subnet then it should already be reachable within that domain, right? So if LB is in the same domain and underlying network is already plumbed I guess I don't need it, right? | 20:45 |
| blogan | manishg: this is the reason i'd like subnet to be optional, if the subnet isn't provided, then the user assumes the member can be connected bc the LB is already on that subnet or L3 | 20:46 |
| blogan | but right now its required | 20:46 |
| blogan | now if a user wants to add a member ona totally different subnet, it does make sense to put the subnet | 20:47 |
| manishg | blogan: thanks. our case is where LB is on L3 and can reach the VM. Trying to understand the usecase of subnet then - so if a VM is on subnetA and LB on another subnet, subnetC. Then there will a shared subnetB in the middle from which the addresses will be allocated to VMs? (and network appropriately plumbed to get those pkts to VM) - not sure if I understood it correctly. | 20:50 |
| rm_work | so when you create an LB | 20:50 |
| rm_work | you get a VIP on a public network segment | 20:50 |
| *** doug-fi__ has joined #openstack-lbaas | 20:50 | |
| rm_work | then when you add members, Octavia plumbs the subnets for those members as well | 20:51 |
| *** doug-fi__ has quit IRC | 20:51 | |
| *** doug-fi__ has joined #openstack-lbaas | 20:51 | |
| blogan | rm_work: doesn't have to be public network segment, coudl be a tenant network | 20:51 |
| rm_work | if you provide 5 members, 3 on user-subnet-A and 2 on user-subnet-B, then it will plumb those two | 20:51 |
| rm_work | ok sure | 20:51 |
| rm_work | fair | 20:51 |
| manishg | rm_work, blogan: thanks. | 20:52 |
| *** doug-fish has quit IRC | 20:53 | |
| rm_work | yeah it really should be optional | 20:54 |
| rm_work | if you for instance don't plan to support private customer subnets | 20:54 |
| rm_work | in that case there is no reason to need to provide a subnet-id because everything is plumbed by default | 20:54 |
| manishg | rm_work: yeah, that's why I was confused about the need of a subnet in the request. But realized the other use-cases after you mentioned it. | 20:56 |
| rm_work | cool, yeah as long as it is clear now :P took me a couple of times too | 20:56 |
| manishg | yeah, if you have tenant subnets, etc. things are little more complex. We use it for large private cloud (all networks shared networks!) - so life is much simpler from networking point of view :) | 20:59 |
| manjeets | blogan: ping | 21:06 |
| manjeets | for creating a listener if protocol value is given in lowercase it return an exception | 21:07 |
| blogan | manjeets: yeah that sounds right, technically, not sure how much of a bad user experience it is | 21:08 |
| *** minwang2 has quit IRC | 21:09 | |
| manjeets | should'nt it be upper and lower case both accepted ? | 21:09 |
| xgerman | ou can always file a bug and we will take it under consideration... | 21:10 |
| blogan | manjeets: perhaps, i would defer to precedence on this one, does security group accept both? | 21:10 |
| *** jongchoi_ has joined #openstack-lbaas | 21:11 | |
| manjeets | it weird in security group ICMP and icmp both are acceptable but http, udp., tcp are only accepted | 21:11 |
| manjeets | except icmp all are lower case | 21:11 |
| manjeets | sorry i just tested in security group TCP,tcp,UDP,udp all are accepted | 21:14 |
| manjeets | blogan: in sec_group both are acccepted | 21:14 |
| blogan | manjeets: okay then yeah we should accept b oth tehn | 21:14 |
| manjeets | i can do that | 21:14 |
| manjeets | it should be converted to uppercase all the time for tests not to break ? | 21:15 |
| *** rtheis has quit IRC | 21:15 | |
| blogan | manjeets: well real question is, does the API accept both lower case and upper case or is the client you're using that does the translation? | 21:16 |
| manjeets | I think Api only accepts uppercase we can convert it in client itself | 21:17 |
| manjeets | I'll check out anyway I filed the bug will fix it if it is in client side or api https://bugs.launchpad.net/neutron/+bug/1508688 | 21:23 |
| openstack | Launchpad bug 1508688 in neutron "lbaasv2- Only uppercase string for protocol are accepted while creating listener" [Undecided,New] | 21:23 |
| *** manjeets has quit IRC | 21:23 | |
| xgerman | mangiest thanks + great for willing to fix | 21:24 |
| xgerman | manjeets | 21:24 |
| *** amotoki has quit IRC | 21:31 | |
| *** doug-fi__ has quit IRC | 21:32 | |
| *** doug-fish has joined #openstack-lbaas | 21:33 | |
| *** manishg has quit IRC | 21:34 | |
| *** doug-fis_ has joined #openstack-lbaas | 21:34 | |
| crc32 | xgerman: can you send me the lab invite for the meeting at 5:30 CST? | 21:37 |
| *** doug-fish has quit IRC | 21:37 | |
| *** doug-fish has joined #openstack-lbaas | 21:38 | |
| *** doug-fis_ has quit IRC | 21:39 | |
| *** doug-fish has quit IRC | 21:42 | |
| *** ducttape_ has quit IRC | 21:49 | |
| *** doug-fish has joined #openstack-lbaas | 21:50 | |
| *** minwang2 has joined #openstack-lbaas | 21:53 | |
| *** doug-fish has quit IRC | 21:54 | |
| *** crc32 has quit IRC | 22:01 | |
| *** jongchoi_ has quit IRC | 22:02 | |
| *** crc32 has joined #openstack-lbaas | 22:03 | |
| *** jongchoi_ has joined #openstack-lbaas | 22:05 | |
| *** jongchoi_ has quit IRC | 22:10 | |
| crc32 | xgerman: can you send me the lab invite for the meeting at 5:30 CST? | 22:19 |
| xgerman | I thought I did carlos.garza@racksoace... | 22:22 |
| xgerman | sent at 2:47 pacific | 22:22 |
| johnsom | Summit hands-on-lab prep walk through in 10 minutes on Google hangout: | 22:23 |
| xgerman | resent | 22:23 |
| johnsom | https://plus.google.com/hangouts/_/hpe.com/summit | 22:23 |
| *** diogogmt has quit IRC | 22:24 | |
| rm_work | xgerman: hangout link? | 22:31 |
| *** fnaval_ has quit IRC | 22:31 | |
| rm_work | it's on my personal calendar not my work one apparently | 22:31 |
| xgerman | https://plus.google.com/hangouts/_/hpe.com/summit | 22:31 |
| rm_work | thanks | 22:31 |
| *** fnaval has joined #openstack-lbaas | 22:35 | |
| openstackgerrit | Merged openstack/neutron-lbaas: Fix default device_driver entry in lbaas_agent.ini https://review.openstack.org/236705 | 22:42 |
| johnsom | https://docs.google.com/presentation/d/132wYD2XsH0uhkotzrgs2ZpH-FMi1OkDEpnxyW1_jnek/edit?usp=sharing | 22:42 |
| crc32 | https://plus.google.com/hangouts/_/hpe.com/summit | 22:50 |
| *** diogogmt has joined #openstack-lbaas | 23:06 | |
| *** diogogmt has quit IRC | 23:12 | |
| *** yamamoto has joined #openstack-lbaas | 23:18 | |
| *** ducttape_ has joined #openstack-lbaas | 23:19 | |
| *** ducttape_ has quit IRC | 23:20 | |
| *** yamamoto has quit IRC | 23:26 | |
| *** amotoki has joined #openstack-lbaas | 23:28 | |
| *** amotoki has quit IRC | 23:33 | |
| *** harlowja_ has quit IRC | 23:33 | |
| *** manjeets has joined #openstack-lbaas | 23:36 | |
| *** harlowja has joined #openstack-lbaas | 23:37 | |
| johnsom | rm_work I could use some python3 help. | 23:39 |
| rm_work | I can try | 23:39 |
| johnsom | python3, import hashlib, import base64 | 23:39 |
| rm_work | currently just watching rsync to USBs | 23:39 |
| johnsom | base64.b64encode(hashlib.sha1(b'foo').digest(), '_-') | 23:39 |
| rm_work | ah is it giving you the encoding error | 23:39 |
| johnsom | TypeError: 'str' does not support the buffer interface | 23:39 |
| rm_work | ah | 23:39 |
| johnsom | Works fine in py27 | 23:39 |
| rm_work | base64.b64encode(hashlib.sha1(b'foo').digest(), b'_-') | 23:40 |
| rm_work | works for me in py3 | 23:40 |
| rm_work | 3.3 and 3.4 | 23:41 |
| johnsom | What? | 23:41 |
| rm_work | only change is b'_-' | 23:41 |
| johnsom | I cut that from my window straight to you | 23:41 |
| johnsom | oh! | 23:41 |
| johnsom | Cool! | 23:41 |
| rm_work | needs to be bytes | 23:41 |
| rm_work | yeah | 23:41 |
| rm_work | basically, bytes ALL THE THINGS | 23:41 |
| johnsom | Thanks! I overlooked that. | 23:41 |
| rm_work | yeah | 23:42 |
| rm_work | easy to do | 23:42 |
| rm_work | took me 1000 times of fixing manual instances | 23:42 |
| rm_work | to really get it :P | 23:42 |
| rm_work | ... since that is basically what i had to do | 23:42 |
| rm_work | or if you want cross-compatible | 23:42 |
| rm_work | you need to use an encoder | 23:42 |
| johnsom | That I hope is the last issue with VRRP and session persistence | 23:42 |
| rm_work | or a decoder | 23:42 |
| rm_work | it's like... | 23:42 |
| *** manjeets has quit IRC | 23:42 | |
| rm_work | str.encode('_-', 'ascii') | 23:43 |
| rm_work | you don't want to put b'string' directly | 23:43 |
| rm_work | in actual polyglot code | 23:43 |
| rm_work | johnsom: ^^ | 23:44 |
| rm_work | and that works on random passed-in args | 23:44 |
| rm_work | just always run stuff through str.encode(var, 'ascii') | 23:44 |
| johnsom | I was using six but boiled down my problem to that code snippet | 23:44 |
| *** crc32 has quit IRC | 23:45 | |
| rm_work | don't need six for most of this | 23:45 |
| rm_work | I believe | 23:45 |
| rm_work | it is a bit overkill | 23:45 |
| *** harlowja_ has joined #openstack-lbaas | 23:45 | |
| johnsom | Show me a fly, I have a sledgehammer | 23:45 |
| xgerman | rm_work they will send an e-mail tonight to the participants of the Octavia lab | 23:48 |
| xgerman | also added Logs + database | 23:48 |
| xgerman | wonder if blogan has some network troubleshooting tips we should share | 23:48 |
| xgerman | ? | 23:48 |
| *** harlowja has quit IRC | 23:49 | |
| *** ducttape_ has joined #openstack-lbaas | 23:54 | |
| *** yamamoto has joined #openstack-lbaas | 23:55 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!