*** lewo has quit IRC | 01:20 | |
*** lewo has joined #openstack-dragonflow | 01:25 | |
*** gongysh has joined #openstack-dragonflow | 01:30 | |
*** yamamoto has quit IRC | 01:45 | |
*** yamamoto has joined #openstack-dragonflow | 01:45 | |
*** kkxue has quit IRC | 02:09 | |
*** kkxue has joined #openstack-dragonflow | 02:09 | |
*** yamamoto has quit IRC | 02:41 | |
*** kkxue has quit IRC | 02:42 | |
*** kkxue has joined #openstack-dragonflow | 02:42 | |
*** yamamoto has joined #openstack-dragonflow | 02:51 | |
*** kkxue has quit IRC | 03:10 | |
*** kkxue has joined #openstack-dragonflow | 03:10 | |
*** yamamoto has quit IRC | 03:33 | |
*** yamamoto has joined #openstack-dragonflow | 03:42 | |
*** yamamoto has quit IRC | 03:48 | |
*** yamamoto has joined #openstack-dragonflow | 03:49 | |
*** yamamoto has quit IRC | 03:54 | |
*** gongysh has quit IRC | 04:15 | |
*** dims has quit IRC | 05:52 | |
*** dims has joined #openstack-dragonflow | 05:54 | |
*** TuanVu has joined #openstack-dragonflow | 06:10 | |
TuanVu | Hi, at this moment my instance is not able to get IP from DHCP | 06:12 |
---|---|---|
TuanVu | as I've checked in OVS, DHCP request is dropped | 06:12 |
TuanVu | could anyone please help to let me know what I could check in this case? | 06:13 |
TuanVu | I've tried to restart df-controller | 06:14 |
kkxue | do you check any log info? | 06:20 |
TuanVu | yes, I've checked log of df-controller, l3 ... | 06:21 |
TuanVu | there's no error | 06:21 |
TuanVu | and also, there's nothing related to dhcp | 06:21 |
TuanVu | the most important point here is: as I've traced in OVS flow table, DHCP request is "dropped" | 06:23 |
TuanVu | as far as I understand, it should be sent from OVS to df-controller and df-controller will process this request | 06:23 |
TuanVu | is there any way to re-generate OVS flow tables? | 06:24 |
kkxue | you can compare this: https://paste.fedoraproject.org/paste/IUeTPObY8M7PRNsKjtr9Nw | 06:28 |
kkxue | Not sure it will helps | 06:29 |
dimak | TuanVu, you're running a devstack? | 06:34 |
TuanVu | yes | 06:34 |
TuanVu | thank you, kkxue | 06:36 |
TuanVu | I'm checking your note | 06:36 |
oanson | TuanVu, yes, you are correct. There should be flows redirecting dhcp to controller. | 06:36 |
dimak | can you share the local.conf you used? | 06:36 |
dimak | Also, make sure DHCP is enabled on the relevant subnet | 06:36 |
oanson | TuanVu, restarting df-controller should regenerate the flows in OVS. | 06:36 |
TuanVu | thanks dimak and oanson for your concern :) | 06:36 |
oanson | TuanVu, don't thank us yet :) | 06:37 |
TuanVu | hi dimak, after many failed troubleshooting attempts, I'm re-installing df on my devstack | 06:38 |
TuanVu | I just want to have a "general guide" for troubleshooting in this case | 06:38 |
TuanVu | like: what should I check, especially for df-controller | 06:39 |
TuanVu | because I think, in this case, there must be something wrong with df-controller, right? | 06:39 |
dimak | I would find out the ovs port the request it received at on br-int | 06:40 |
dimak | then try tracing it | 06:40 |
TuanVu | Hi oanson, oh thanks for this information | 06:40 |
TuanVu | @dimak: yes, I agree | 06:40 |
TuanVu | that's what I did and found out DHCP request is dropped | 06:41 |
oanson | TuanVu, we have started a guide here: https://docs.openstack.org/dragonflow/latest/testing_and_debugging.html but it is in early stages :( | 06:41 |
dimak | Where it is dropped? | 06:41 |
TuanVu | @oanson: yeah, I've also checked that | 06:41 |
TuanVu | @dimak: this traffic is forwareded from one table to another for serveral time (5 > 10 > 15) | 06:43 |
TuanVu | at table 15, it's dropped and I have no idea why | 06:43 |
dimak | TuanVu, I'll compare to my env in a moment, just need 5 minutes to grab a coffee | 06:43 |
TuanVu | I mean ... how to find out "why it is dropped", is there any way to check? | 06:44 |
oanson | TuanVu, could you upload your OVS flows? | 06:44 |
TuanVu | thank you @dimak, that's really nice of you :) | 06:44 |
TuanVu | but don't worry about it too much, I'm in the process of re-installing df in my lab anyway | 06:44 |
TuanVu | @oanson: I'll provide more details next time | 06:45 |
TuanVu | @all: thank you guys for your awesome support :) | 06:46 |
TuanVu | I really appreciate it | 06:46 |
oanson | TuanVu, no worries. Thanks for trying Dragonflow out | 06:46 |
kkxue | dimak, I've tested your SFC heat template,and success build the test environment. | 06:50 |
TuanVu | Hi oanson, Dragonflow is a really cool project. I really like it | 06:51 |
TuanVu | I'm trying to understand more about dataflow and will read the source code for more details | 06:51 |
TuanVu | thank you guys for your great work | 06:51 |
oanson | Thanks :) | 06:52 |
dimak | kkxue, thanks! | 07:00 |
kkxue | but $server_fip test failed.. | 07:01 |
kkxue | vms/networks/pp/ppg/fc have created. | 07:03 |
dimak | kkxue, you were not able to get the echo back? | 07:06 |
kkxue | yes | 07:06 |
dimak | Did you wait a couple of minutes? I takes ovs+ryu a few moments to install | 07:07 |
dimak | if you still have the stack up, i'd like to see some log file | 07:08 |
kkxue | which log file do you want to see? | 07:09 |
dimak | /var/log/cloud-init-output.log on sfc-test-sf VM | 07:11 |
kkxue | password of Fedora-Cloud-Base-25-1.3.x86_64 is? | 07:12 |
kkxue | I haven't use this image before. | 07:13 |
dimak | it should use the keypair | 07:13 |
dimak | login to user fedora | 07:13 |
dimak | If you run 'openstack server list' you'll see it has a fip | 07:14 |
kkxue | yes | 07:16 |
*** NatanBro has joined #openstack-dragonflow | 07:17 | |
kkxue | dimak,/var/log/cloud-init-output.log show error: "/tmp/run.sh: line 11: ryu-manager-3: command not found" | 07:33 |
dimak | kkxue, did dnf install succeed? | 07:34 |
kkxue | oh,my pulic network have no internet access.. | 07:34 |
dimak | oh.. | 07:36 |
dimak | that's a deal breaker, i need to install ovs and ryu to implement the service function | 07:36 |
oanson | kkxue, maybe you can set up a local dnf mirror? | 07:39 |
kkxue | I will change the public network | 07:41 |
kkxue | btw i found a bug that redis-server service have not shutdown after unstack.sh | 07:42 |
kkxue | if you use redis database. | 07:44 |
oanson | kkxue, I think it's reported here: https://bugs.launchpad.net/dragonflow/+bug/1706862 . Try killall -9 redis-server between unstack and stack. | 07:45 |
openstack | Launchpad bug 1706862 in DragonFlow "re-stack with redis failed. unstack doesn't kill redis-server" [Low,New] | 07:45 |
kkxue | yes,that's it | 07:45 |
kkxue | dimak,maybe you should notice this to your template file. | 07:49 |
dimak | Ok, I will add, thanks! | 07:51 |
kkxue | and why not use a open-source vnf image? | 07:52 |
kkxue | how traker project do tests like this/ | 07:55 |
irenab | oanson, hi | 07:55 |
oanson | irenab, hi | 07:55 |
irenab | oanson, tired of PTO? | 07:55 |
oanson | I can do both. Cooperative threading! :) | 07:56 |
irenab | oanson, once you have few mins, would like to chat about integration with 3rd party LBaaS providers | 07:57 |
irenab | seems to be broken/not supported now when L3agent is disabled | 07:57 |
oanson | Sure. We can do that now | 07:57 |
irenab | dimak, do you have few mins ^^ | 07:58 |
irenab | https://bugs.launchpad.net/dragonflow/+bug/1708178 | 07:59 |
openstack | Launchpad bug 1708178 in DragonFlow "LBaaSv2 with 3rd party provider does not work if L3agent is disabled" [Undecided,New] | 07:59 |
dimak | irenab, yes | 07:59 |
oanson | Yes | 08:00 |
irenab | first, please set the priority for this bug | 08:00 |
dimak | I think we'll need to come up with an lb app that forwards packets bound to VIP port into the relevant compute/namespace | 08:00 |
oanson | The truth is I didn't know we support this feature :) | 08:00 |
oanson | irenab, what priority would you like? High? | 08:00 |
irenab | oanson, it worked before (with l3), this is how kuryr implements k8s services | 08:01 |
irenab | oanson, for kuryr integration its critical | 08:01 |
oanson | l3 agent is still supported. So that can be a workaround until this bug is fixed | 08:01 |
oanson | It can also be gated in the Kuryr-Dragonflow integration gate (which we can add to Dragonflow gate), so it will even be gated | 08:02 |
dimak | dnat acts funny with l3 | 08:02 |
irenab | I think now with l3 agent we have some conflicts with DNAT, dimak has much more details | 08:02 |
dimak | because both will plug into br-ex and answer arps | 08:02 |
oanson | I'm not sure I want to say we don't support l3 agent. | 08:02 |
oanson | Is there a way we can get to play nice? | 08:02 |
oanson | Unless we do better on all fronts. Then we can do away with it completely. | 08:03 |
oanson | (And then we might do well to remove it from our code base) | 08:03 |
irenab | oanson, I agree with your approach | 08:03 |
oanson | Do we know if we are feature-compatible? | 08:04 |
irenab | lets keep L3agent enabled deployment possible, since it seems to be enabler for 3rd party services integ for now | 08:05 |
irenab | oanson, if DF supports all l3 extensions, I beleive the answer is yes | 08:05 |
dimak | I'll see if l3+ no dnat app works finne | 08:05 |
dimak | fine* | 08:05 |
irenab | is it possible to disable DNAT on l3agent? | 08:06 |
oanson | Is it possible to support l3 extensions in general, or do we have to support one by one? e.g. a single l3 extensions app, or do we need an app for lbaas, and one for e.g. vpn, etc? | 08:06 |
oanson | (I don't remember the API. Someone will have to look it up if we don't know) | 08:07 |
irenab | lbaas, vpnaas, fwaas are not l3 extensiosn | 08:07 |
irenab | each service is a separate extension set, and may have even separate api server (lbaas) is now octavia and not neutron | 08:07 |
oanson | Then l3 extensions is not enough. | 08:07 |
irenab | but they have proper integration into neutron by creating nets, ports, ets | 08:08 |
irenab | etc | 08:08 |
oanson | Sounds like we need to keep l3 agent support, since new extensions may be written to first support it, and if Dragonflow supports it we get it for free | 08:09 |
irenab | oanson, I will try to summarize | 08:10 |
irenab | 1. Keep L3agent support in DF (+fixing the issues to make it work again) | 08:11 |
irenab | 2. Advanced net services, lbaas (and maybe vpn) made working in l3agent enabled deployment | 08:12 |
irenab | 3. Provide DF native advanced services support | 08:12 |
irenab | correct? | 08:12 |
oanson | This is a task-list, or a multiple choice? | 08:12 |
oanson | I guess we want 1 and 3, and let the deployer select what they need. | 08:13 |
irenab | more of ma task list, or plan | 08:13 |
oanson | I figured 2 is a subset of these | 08:13 |
irenab | for 2, we need to make sure it works | 08:13 |
oanson | We prefer DF native services when possible - then we are fully distributed and can reach the scale we want. But realistically, we aren't enough people to keep up with OpenStack. | 08:14 |
irenab | and be able to integrate with 3rd party services proveders is good | 08:14 |
oanson | So yes, I guess we need all the above, and I guess in order, since it will take a while before we will be first-to-market on someone else's advanced network service. | 08:14 |
oanson | Yes, that's also true. | 08:15 |
oanson | So I guess we're in agreement. | 08:15 |
dimak | Well | 08:15 |
oanson | Item 1 and 2 are critical, sadly. | 08:16 |
dimak | Once we'll have DF l3 flavor we'll be able to use both df l3 and l3 agent | 08:16 |
dimak | by flavor I mean neutron side service provider | 08:16 |
oanson | Yes, but that might take time. | 08:16 |
oanson | I don't want it to block kuryr integration if there's a choice | 08:17 |
dimak | Yes | 08:17 |
dimak | maybe we can split our gate jobs | 08:17 |
dimak | to test both l3 app and l3 agent | 08:17 |
oanson | We will have to - this needs to be gated. | 08:17 |
irenab | agree | 08:18 |
oanson | But maybe we can gate the l3 agent with kuryr integration | 08:18 |
*** openstackgerrit has quit IRC | 08:18 | |
irenab | in kuryr or df ? | 08:18 |
oanson | This way we also have a real-live scenario of where this is used. | 08:18 |
oanson | irenab, in df. | 08:18 |
irenab | for now we only have deployment gate in kuryr, so this can be great | 08:18 |
oanson | We want to find bugs in Dragonflow and prevent them. | 08:18 |
irenab | oanson, +1 | 08:19 |
oanson | If kuryr runs a kuryr-df integration gate to verify kuryr doesn't break Dragonflow, I'm all for it. But I don't expect a different project to gate for our bugs | 08:19 |
irenab | oanson, I am alligned with you | 08:20 |
oanson | irenab, is the integration far enough along to create a non-voting gate? Even if it fails a lot? | 08:20 |
irenab | oanson, speaking of kuryr I am not sure, sicne there is a lot of work dmellado is doing | 08:21 |
irenab | for the testing at the gate | 08:22 |
irenab | on DF side, we will need deployment to include kuryr instead or in addition to nova | 08:22 |
oanson | Like we have for bgp and sfc? | 08:23 |
irenab | probably | 08:23 |
oanson | Doesn't sound like an issue. | 08:23 |
irenab | kury is sort of replacement to novawith regards to container | 08:23 |
irenab | oanson, dimak : cool. Sounds like there is a plan | 08:24 |
oanson | Yes. | 08:24 |
oanson | dimak, irenab, if you need an extra pair of hands, enlist leyal. I think he is working on bugs now, and we said these are critical. | 08:24 |
dimak | sure | 08:24 |
irenab | oanson, thanks | 08:25 |
oanson | Setting the bug to critical, and adding this log. | 08:25 |
dimak | I'll try to stabilize l3 app + no agent + l3 agent + no dnat | 08:25 |
irenab | oanson, have a nice rest of the PTO | 08:25 |
oanson | dimak, will this also solve the critical dnat bug? | 08:25 |
oanson | irenab, thanks :) | 08:25 |
oanson | dimak, irenab, leyal, I didn't add an assignee. I'm guessing the three of you will work it out. | 08:27 |
oanson | Thanks! | 08:27 |
dimak | +1 | 08:27 |
irenab | oanson, I am on PTO (for real) next week | 08:27 |
oanson | My PTO is real! But I hope you enjoy yours more :) | 08:28 |
oanson | Will you be back the week after? | 08:28 |
irenab | hope so | 08:28 |
leyal | dimak, irenab - which bugs we are talking about? \ | 08:29 |
oanson | leyal, https://bugs.launchpad.net/dragonflow/+bug/1708178 | 08:29 |
openstack | Launchpad bug 1708178 in DragonFlow "LBaaSv2 with 3rd party provider does not work if L3agent is disabled" [Critical,New] | 08:29 |
leyal | just saw now the long discussion in the past half an hour .. i will read it .. | 08:31 |
*** NatanBro has quit IRC | 08:31 | |
oanson | No worries. :) | 08:31 |
*** zenoway has joined #openstack-dragonflow | 08:32 | |
*** zenoway has quit IRC | 08:35 | |
*** zenoway has joined #openstack-dragonflow | 08:36 | |
*** dims has quit IRC | 09:04 | |
*** dims has joined #openstack-dragonflow | 09:08 | |
*** zenoway has quit IRC | 09:30 | |
*** zenoway has joined #openstack-dragonflow | 09:31 | |
*** yamamoto has joined #openstack-dragonflow | 09:41 | |
*** zenoway has quit IRC | 09:44 | |
*** kkxue_ has joined #openstack-dragonflow | 09:53 | |
*** kkxue has quit IRC | 09:56 | |
*** TuanVu has quit IRC | 10:06 | |
*** zenoway has joined #openstack-dragonflow | 10:09 | |
*** kkxue_ has quit IRC | 10:44 | |
*** kkxue has joined #openstack-dragonflow | 11:47 | |
*** kkxue has quit IRC | 11:52 | |
*** kkxue has joined #openstack-dragonflow | 11:52 | |
*** openstackgerrit has joined #openstack-dragonflow | 12:12 | |
openstackgerrit | Dima Kuznetsov proposed openstack/dragonflow master: Devstack: fix redis unstack shutdown https://review.openstack.org/490452 | 12:12 |
openstackgerrit | Dima Kuznetsov proposed openstack/dragonflow master: Devstack: fix redis unstack shutdown https://review.openstack.org/490452 | 12:21 |
*** kkxue has quit IRC | 12:42 | |
*** kkxue has joined #openstack-dragonflow | 12:42 | |
*** kkxue has quit IRC | 12:44 | |
openstackgerrit | Dima Kuznetsov proposed openstack/dragonflow master: Move tempest tests filter to a common place https://review.openstack.org/489002 | 12:44 |
openstackgerrit | Dima Kuznetsov proposed openstack/dragonflow master: [DNM][WIP] Fix tempest https://review.openstack.org/489003 | 12:44 |
openstackgerrit | Dima Kuznetsov proposed openstack/dragonflow master: Disable l3 agent in gate https://review.openstack.org/483385 | 12:44 |
openstackgerrit | Dima Kuznetsov proposed openstack/dragonflow master: DNAT: handle port creation correctly https://review.openstack.org/489000 | 12:55 |
dimak | irenab, posted reply on https://review.openstack.org/#/c/489000/ | 12:56 |
*** yamamoto has quit IRC | 13:06 | |
*** natanb has joined #openstack-dragonflow | 13:18 | |
*** mlavalle has joined #openstack-dragonflow | 13:58 | |
*** yamamoto has joined #openstack-dragonflow | 14:06 | |
irenab | dimak: will check shortly | 14:09 |
openstackgerrit | Dima Kuznetsov proposed openstack/dragonflow master: [DNM][WIP] Fix tempest https://review.openstack.org/489003 | 14:11 |
*** yamamoto has quit IRC | 14:11 | |
*** irenab has left #openstack-dragonflow | 14:14 | |
openstackgerrit | Lihi Wishnitzer proposed openstack/dragonflow master: [WIP] Switch etcd base driver to use etcd3gw https://review.openstack.org/489246 | 14:24 |
*** zenoway has quit IRC | 14:32 | |
*** zenoway has joined #openstack-dragonflow | 14:33 | |
*** irenab has joined #openstack-dragonflow | 14:41 | |
*** zenoway has quit IRC | 14:42 | |
*** yamamoto has joined #openstack-dragonflow | 14:47 | |
*** yamamoto has quit IRC | 14:47 | |
*** natanb has quit IRC | 14:49 | |
*** zenoway has joined #openstack-dragonflow | 14:52 | |
*** zenoway has quit IRC | 14:58 | |
*** zenoway has joined #openstack-dragonflow | 14:58 | |
*** zenoway has quit IRC | 15:02 | |
*** zenoway has joined #openstack-dragonflow | 15:02 | |
*** zenoway has quit IRC | 15:07 | |
openstackgerrit | Eyal Leshem proposed openstack/dragonflow master: Security-groups - support also number as ip-protocol https://review.openstack.org/489007 | 15:13 |
openstackgerrit | Eyal Leshem proposed openstack/dragonflow master: Security-groups - support also number as ip-protocol https://review.openstack.org/489007 | 15:15 |
*** yamamoto has joined #openstack-dragonflow | 15:48 | |
*** yamamoto has quit IRC | 15:53 | |
*** yamamoto has joined #openstack-dragonflow | 21:21 | |
*** yamamoto has quit IRC | 21:26 | |
*** yamamoto has joined #openstack-dragonflow | 21:58 | |
*** mlavalle has quit IRC | 23:41 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!