| *** enikanorov__ has joined #openstack-neutron | 00:00 | |
| marun | nati_ueno: so, how important is allowed address pair? | 00:00 |
|---|---|---|
| nati_ueno | marun: https://blueprints.launchpad.net/neutron/+spec/allowed-address-pairs | 00:00 |
| nati_ueno | marun: I'm not sure how important it is. However havana is already shipped | 00:00 |
| marun | nati_ueno: maybe we could use nwfilter in nova for havana and do it properly for icehouse? | 00:01 |
| nati_ueno | marun: so we can't remove it by policy | 00:01 |
| marun | nati_ueno: or give people an option for havana? | 00:01 |
| marun | nati_ueno: choose - allowed address pair or arp spoofing protection | 00:01 |
| nati_ueno | marun: I think that's good start | 00:01 |
| marun | nati_ueno: I'll respond to the email with as much... | 00:02 |
| nati_ueno | marun: Thanks | 00:02 |
| *** enikanorov_ has quit IRC | 00:03 | |
| *** mlavalle has quit IRC | 00:04 | |
| openstackgerrit | Mohammad Banikazemi proposed a change to openstack/neutron: Deals with fails in update-*-postcommit ops https://review.openstack.org/69792 | 00:06 |
| openstackgerrit | Mohammad Banikazemi proposed a change to openstack/neutron: Adds the new IBM SDN-VE plugin https://review.openstack.org/66453 | 00:10 |
| *** thuc has quit IRC | 00:10 | |
| *** banix has quit IRC | 00:11 | |
| *** thuc has joined #openstack-neutron | 00:11 | |
| *** nati_ueno has quit IRC | 00:11 | |
| *** salparadise has joined #openstack-neutron | 00:13 | |
| *** thuc has quit IRC | 00:15 | |
| *** nati_ueno has joined #openstack-neutron | 00:19 | |
| *** nati_ueno has quit IRC | 00:22 | |
| *** zzelle has quit IRC | 00:23 | |
| *** nati_ueno has joined #openstack-neutron | 00:23 | |
| *** matsuhashi has joined #openstack-neutron | 00:24 | |
| *** thedodd has quit IRC | 00:29 | |
| *** dims has quit IRC | 00:31 | |
| *** dims has joined #openstack-neutron | 00:32 | |
| *** shivh has quit IRC | 00:32 | |
| *** jhurlbert has quit IRC | 00:35 | |
| *** sn6i23a has joined #openstack-neutron | 00:43 | |
| *** IanGovett has quit IRC | 00:45 | |
| *** sbalukoff has quit IRC | 00:46 | |
| openstackgerrit | Rajesh Mohan proposed a change to openstack/neutron: Agent implementation for SSL-VPN services https://review.openstack.org/70274 | 00:47 |
| *** aymenfrikha has joined #openstack-neutron | 00:49 | |
| *** pheadron1 has joined #openstack-neutron | 00:51 | |
| *** pheadron has quit IRC | 00:54 | |
| *** yamahata has quit IRC | 00:59 | |
| *** evgenyf has joined #openstack-neutron | 01:01 | |
| *** sputnik13net has quit IRC | 01:01 | |
| *** markmcclain has quit IRC | 01:01 | |
| *** markmcclain has joined #openstack-neutron | 01:02 | |
| *** MM_at_HP has quit IRC | 01:03 | |
| *** harlowja is now known as harlowja_away | 01:04 | |
| *** sputnik13net has joined #openstack-neutron | 01:04 | |
| *** shivh has joined #openstack-neutron | 01:11 | |
| *** nati_uen_ has joined #openstack-neutron | 01:11 | |
| openstackgerrit | Nachi Ueno proposed a change to openstack/neutron: Agent implementation for SSL-VPN services https://review.openstack.org/70274 | 01:13 |
| nati_uen_ | anteaya: Hi | 01:14 |
| *** nati_ueno has quit IRC | 01:14 | |
| *** banix has joined #openstack-neutron | 01:19 | |
| openstackgerrit | Mohammad Banikazemi proposed a change to openstack/neutron: Deals with fails in update-*-postcommit ops https://review.openstack.org/69792 | 01:26 |
| *** harlowja_away is now known as harlowja | 01:26 | |
| *** IanGovett has joined #openstack-neutron | 01:29 | |
| *** salv-orlando has quit IRC | 01:30 | |
| *** evgenyf has quit IRC | 01:30 | |
| *** sputnik13net has quit IRC | 01:32 | |
| openstackgerrit | Nachi Ueno proposed a change to openstack/neutron: Agent implementation for SSL-VPN services https://review.openstack.org/70274 | 01:34 |
| *** nati_uen_ has quit IRC | 01:36 | |
| *** banix has quit IRC | 01:38 | |
| *** markmcclain has quit IRC | 01:40 | |
| *** terryw has quit IRC | 01:41 | |
| *** banix has joined #openstack-neutron | 01:46 | |
| *** IanGovett has quit IRC | 01:54 | |
| *** networkstatic has joined #openstack-neutron | 01:55 | |
| *** pheadron1 has quit IRC | 01:56 | |
| *** sn6i23a has quit IRC | 02:00 | |
| *** markwash has quit IRC | 02:02 | |
| *** markwash has joined #openstack-neutron | 02:03 | |
| *** yamahata has joined #openstack-neutron | 02:03 | |
| *** terryw has joined #openstack-neutron | 02:04 | |
| *** banix has quit IRC | 02:09 | |
| *** networkstatic has quit IRC | 02:16 | |
| *** networkstatic has joined #openstack-neutron | 02:18 | |
| *** networkstatic has quit IRC | 02:18 | |
| *** WackoRobie has joined #openstack-neutron | 02:18 | |
| *** thedodd has joined #openstack-neutron | 02:18 | |
| *** networkstatic has joined #openstack-neutron | 02:18 | |
| *** networkstatic has quit IRC | 02:19 | |
| *** networkstatic has joined #openstack-neutron | 02:19 | |
| *** giulivo has quit IRC | 02:20 | |
| *** shashank_ has quit IRC | 02:30 | |
| *** godara has quit IRC | 02:31 | |
| *** aymenfrikha has quit IRC | 02:31 | |
| *** matsuhashi has quit IRC | 02:34 | |
| *** shivh has quit IRC | 02:35 | |
| *** matsuhas_ has joined #openstack-neutron | 02:37 | |
| *** thedodd has quit IRC | 02:44 | |
| *** networkstatic has quit IRC | 02:44 | |
| *** gdubreui has quit IRC | 03:04 | |
| *** dave_tucker is now known as dave_tucker_zzz | 03:06 | |
| *** matsuhashi has joined #openstack-neutron | 03:07 | |
| *** shashank_ has joined #openstack-neutron | 03:07 | |
| *** matsuhas_ has quit IRC | 03:08 | |
| *** matsuhashi has quit IRC | 03:08 | |
| *** shashank_ has quit IRC | 03:11 | |
| openstackgerrit | Yuuichi Fujioka proposed a change to openstack/neutron: fix to allow setting ip address of floating ip https://review.openstack.org/70286 | 03:15 |
| *** emagana has quit IRC | 03:16 | |
| *** matsuhashi has joined #openstack-neutron | 03:20 | |
| *** matsuhashi has quit IRC | 03:34 | |
| *** matsuhashi has joined #openstack-neutron | 03:35 | |
| *** thedodd has joined #openstack-neutron | 03:38 | |
| openstackgerrit | Mohammad Banikazemi proposed a change to openstack/neutron: Deals with fails in update-*-postcommit ops https://review.openstack.org/69792 | 03:39 |
| *** markwash has quit IRC | 03:41 | |
| *** matsuhashi has quit IRC | 03:43 | |
| *** sweston has quit IRC | 03:43 | |
| *** matsuhashi has joined #openstack-neutron | 03:50 | |
| *** sweston has joined #openstack-neutron | 03:52 | |
| *** sweston has quit IRC | 03:55 | |
| *** sweston has joined #openstack-neutron | 03:57 | |
| *** sn6i23a has joined #openstack-neutron | 04:06 | |
| *** changbl has joined #openstack-neutron | 04:07 | |
| *** WackoRobie has quit IRC | 04:12 | |
| *** HenryG has joined #openstack-neutron | 04:13 | |
| *** thedodd has quit IRC | 04:18 | |
| openstackgerrit | oda-g proposed a change to openstack/neutron: Enhance GET networks performance of metaplugin https://review.openstack.org/70060 | 04:19 |
| openstackgerrit | oda-g proposed a change to openstack/neutron: Enhance GET networks performance of metaplugin https://review.openstack.org/70060 | 04:22 |
| *** harlowja is now known as harlowja_away | 04:26 | |
| openstackgerrit | Yuuichi Fujioka proposed a change to openstack/neutron: fix to allow setting ip address of floating ip https://review.openstack.org/70286 | 04:27 |
| *** markwash has joined #openstack-neutron | 04:32 | |
| HenryG | rkukura: ping? | 04:45 |
| *** tongli has quit IRC | 04:49 | |
| *** markwash has quit IRC | 04:53 | |
| *** markwash has joined #openstack-neutron | 04:56 | |
| *** Mierdin has quit IRC | 04:59 | |
| *** markwash has quit IRC | 05:03 | |
| openstackgerrit | Mohammad Banikazemi proposed a change to openstack/neutron: Deals with fails in update-*-postcommit ops https://review.openstack.org/69792 | 05:13 |
| *** WackoRobie has joined #openstack-neutron | 05:22 | |
| *** vkozhukalov has joined #openstack-neutron | 05:25 | |
| *** WackoRobie has quit IRC | 05:27 | |
| *** HenryG has quit IRC | 05:32 | |
| *** sn6i23a has quit IRC | 05:33 | |
| *** amotoki has joined #openstack-neutron | 05:37 | |
| *** thuc has joined #openstack-neutron | 05:41 | |
| *** shausy has joined #openstack-neutron | 05:42 | |
| *** thuc_ has joined #openstack-neutron | 05:42 | |
| *** thuc has quit IRC | 05:46 | |
| *** pradipta has joined #openstack-neutron | 05:47 | |
| *** simon-AS559 has joined #openstack-neutron | 05:52 | |
| *** simon-AS559 has quit IRC | 05:58 | |
| *** simon-AS559 has joined #openstack-neutron | 05:58 | |
| *** rwsu has quit IRC | 05:59 | |
| *** afazekas_ has joined #openstack-neutron | 06:15 | |
| *** afazekas_ has quit IRC | 06:15 | |
| enikanorov__ | mestery: hi | 06:17 |
| enikanorov__ | mestery: got this patch working https://review.openstack.org/#/c/66670/ i think we can let it in | 06:17 |
| enikanorov__ | for some reason it didn't hit kernel crash 3 times in a row | 06:18 |
| openstackgerrit | Jenkins proposed a change to openstack/neutron: Imported Translations from Transifex https://review.openstack.org/66501 | 06:24 |
| *** jprovazn has joined #openstack-neutron | 06:25 | |
| *** mflobo_ has joined #openstack-neutron | 06:27 | |
| *** mflobo has quit IRC | 06:30 | |
| *** mflobo has joined #openstack-neutron | 06:33 | |
| *** anand has joined #openstack-neutron | 06:35 | |
| *** mflobo_ has quit IRC | 06:35 | |
| *** shausy has quit IRC | 06:36 | |
| *** shausy has joined #openstack-neutron | 06:37 | |
| *** terryw has quit IRC | 06:38 | |
| *** alex_klimov has joined #openstack-neutron | 06:42 | |
| *** thuc_ has quit IRC | 06:47 | |
| *** thuc has joined #openstack-neutron | 06:47 | |
| *** emagana has joined #openstack-neutron | 06:48 | |
| *** vkozhukalov has quit IRC | 06:51 | |
| *** thuc has quit IRC | 06:52 | |
| *** bashok has joined #openstack-neutron | 07:06 | |
| *** shashank_ has joined #openstack-neutron | 07:11 | |
| *** sbalukoff has joined #openstack-neutron | 07:15 | |
| *** simon-AS559 has quit IRC | 07:22 | |
| *** simon-AS559 has joined #openstack-neutron | 07:23 | |
| *** djoreilly has joined #openstack-neutron | 07:25 | |
| *** simon-AS559 has quit IRC | 07:27 | |
| *** humbolt has joined #openstack-neutron | 07:28 | |
| *** humbolt1 has quit IRC | 07:30 | |
| *** rohit404 has joined #openstack-neutron | 07:41 | |
| *** shausy has quit IRC | 07:43 | |
| *** shausy has joined #openstack-neutron | 07:44 | |
| *** markwash has joined #openstack-neutron | 07:46 | |
| *** dguitarbite has quit IRC | 07:55 | |
| *** markwash has quit IRC | 07:59 | |
| *** markwash has joined #openstack-neutron | 08:04 | |
| *** amotoki_ has joined #openstack-neutron | 08:07 | |
| openstackgerrit | Yohei Matsuhashi proposed a change to openstack/python-neutronclient: Enable to select specific network service type https://review.openstack.org/54534 | 08:09 |
| *** markwash has quit IRC | 08:11 | |
| *** simon-AS559 has joined #openstack-neutron | 08:16 | |
| *** simon-AS5591 has joined #openstack-neutron | 08:19 | |
| *** simon-AS559 has quit IRC | 08:20 | |
| *** vkozhukalov has joined #openstack-neutron | 08:29 | |
| *** matsuhashi has quit IRC | 08:37 | |
| *** amotoki_ has quit IRC | 08:37 | |
| openstackgerrit | Édouard Thuleau proposed a change to openstack/neutron: Implement local ARP responder onto OVS agent https://review.openstack.org/49227 | 08:39 |
| *** ygbo has joined #openstack-neutron | 08:42 | |
| openstackgerrit | Édouard Thuleau proposed a change to openstack/neutron: Implement local ARP responder onto OVS agent https://review.openstack.org/49227 | 08:43 |
| *** AlexF_ has joined #openstack-neutron | 08:47 | |
| *** alexpilotti has joined #openstack-neutron | 08:47 | |
| *** jpich has joined #openstack-neutron | 08:50 | |
| openstackgerrit | Édouard Thuleau proposed a change to openstack/neutron: OVS lib defer apply doesn't handle concurrency https://review.openstack.org/63917 | 08:51 |
| *** alexpilotti has quit IRC | 08:52 | |
| *** jistr has joined #openstack-neutron | 08:58 | |
| *** safchain has joined #openstack-neutron | 09:08 | |
| *** shashank_ has quit IRC | 09:10 | |
| *** AlexF_ has quit IRC | 09:17 | |
| *** jlibosva has joined #openstack-neutron | 09:29 | |
| *** amritanshu_RnD has joined #openstack-neutron | 09:40 | |
| *** amritanshu_RnD is now known as Guest6335 | 09:40 | |
| *** matsuhashi has joined #openstack-neutron | 09:44 | |
| *** changbl has quit IRC | 09:49 | |
| openstackgerrit | Aleks Chirko proposed a change to openstack/neutron: Bugfix and refactoring for ovs_lib flow mgnt methods https://review.openstack.org/58533 | 09:50 |
| *** jp_at_hp has joined #openstack-neutron | 09:55 | |
| *** rossella_s has joined #openstack-neutron | 10:04 | |
| *** matsuhashi has quit IRC | 10:05 | |
| openstackgerrit | Sylvain Afchain proposed a change to openstack/neutron: L3 Metering label as shared https://review.openstack.org/70090 | 10:05 |
| *** WackoRobie has joined #openstack-neutron | 10:07 | |
| *** rossella_s has quit IRC | 10:10 | |
| *** matsuhashi has joined #openstack-neutron | 10:16 | |
| openstackgerrit | Sylvain Afchain proposed a change to openstack/neutron: Allow multiple DNS forwarders for dnsmasq https://review.openstack.org/62002 | 10:26 |
| *** ykaneko has joined #openstack-neutron | 10:38 | |
| *** Jabadia has joined #openstack-neutron | 10:45 | |
| *** Jabadia has quit IRC | 10:46 | |
| *** Jabadia has joined #openstack-neutron | 10:46 | |
| *** rossella_s has joined #openstack-neutron | 10:56 | |
| ygbo | Hi anteaya, can I "recheck no bug" on https://review.openstack.org/#/c/52930/ or is there still some test issues? | 11:00 |
| *** ykaneko has quit IRC | 11:02 | |
| *** WackoRobie has quit IRC | 11:08 | |
| *** WackoRobie has joined #openstack-neutron | 11:09 | |
| *** Mierdin has joined #openstack-neutron | 11:13 | |
| *** WackoRobie has quit IRC | 11:13 | |
| *** shausy has quit IRC | 11:33 | |
| *** shausy has joined #openstack-neutron | 11:34 | |
| *** IanGovett has joined #openstack-neutron | 11:40 | |
| *** IanGovett1 has joined #openstack-neutron | 11:41 | |
| *** IanGovett has quit IRC | 11:44 | |
| openstackgerrit | András Gyácsok proposed a change to openstack/neutron: Removed copyright from empty files https://review.openstack.org/63901 | 11:48 |
| *** emagana has quit IRC | 11:48 | |
| anand | hello all, why my cirros vm is not getting any internal IP. from dashboard I can see an IP is assigned to vm. but not inside vm | 11:49 |
| anand | pls find the log file http://paste.openstack.org/show/62232/ | 11:50 |
| *** pcm has joined #openstack-neutron | 11:51 | |
| *** shausy has quit IRC | 11:51 | |
| *** pcm has quit IRC | 11:53 | |
| *** morganfainberg is now known as morganfainberg|z | 11:53 | |
| *** pcm has joined #openstack-neutron | 11:53 | |
| *** bashok has quit IRC | 11:56 | |
| *** Jabadia has quit IRC | 12:07 | |
| *** HenryG has joined #openstack-neutron | 12:14 | |
| *** WackoRobie has joined #openstack-neutron | 12:19 | |
| *** bada_ has joined #openstack-neutron | 12:21 | |
| openstackgerrit | Isaku Yamahata proposed a change to openstack/neutron: Implement service vm framework(WIP): https://review.openstack.org/56892 | 12:23 |
| *** bada has quit IRC | 12:24 | |
| *** WackoRobie has quit IRC | 12:24 | |
| *** Jabadia has joined #openstack-neutron | 12:38 | |
| *** Jabadia has quit IRC | 12:42 | |
| *** emagana has joined #openstack-neutron | 12:49 | |
| jlibosva | haleyb: Hi Brian | 12:52 |
| *** yamahata has quit IRC | 12:54 | |
| *** emagana has quit IRC | 12:58 | |
| *** julim has joined #openstack-neutron | 13:00 | |
| *** markvoelker has joined #openstack-neutron | 13:07 | |
| *** aburaschi has joined #openstack-neutron | 13:08 | |
| *** dave_tucker_zzz is now known as dave_tucker | 13:08 | |
| *** dave_tucker is now known as dave_tucker_zzz | 13:13 | |
| *** markvoelker has quit IRC | 13:14 | |
| *** dave_tucker_zzz is now known as dave_tucker | 13:18 | |
| *** markvoelker has joined #openstack-neutron | 13:19 | |
| *** WackoRob_ has joined #openstack-neutron | 13:20 | |
| *** aymenfrikha has joined #openstack-neutron | 13:28 | |
| rkukura | HenryG: pong | 13:37 |
| *** WackoRob_ has quit IRC | 13:38 | |
| HenryG | rkukura: Good morning. Just wanted to alert you to a BP for a new ML2 mech driver. https://blueprints.launchpad.net/neutron/+spec/ml2-cisco-apic-mechanism-driver | 13:38 |
| *** Jabadia has joined #openstack-neutron | 13:39 | |
| HenryG | rkukura: Hoping to get it approved for I-3. | 13:39 |
| rkukura | HenryG: Cool. Will take care of it. | 13:40 |
| HenryG | rkukura: thanks! | 13:40 |
| *** matsuhashi has quit IRC | 13:41 | |
| rkukura | HenryG: Couple of questions after a quick read of the spec: | 13:42 |
| HenryG | rkukura: sure ... | 13:42 |
| *** Jabadia has quit IRC | 13:43 | |
| rkukura | HenryG: 1) Does this MD does need to be involved in port tbinding, or is that handled by the existing openvswitch-agent MD? | 13:44 |
| HenryG | rkukura: the latter | 13:44 |
| *** baoli has joined #openstack-neutron | 13:45 | |
| rkukura | HenryG: 2) From the openvswitch MD's point of view, are these VLAN networks? | 13:45 |
| HenryG | rkukura: yes | 13:45 |
| HenryG | rkukura: we might support vxlan too, or that might slip to juno | 13:47 |
| rkukura | HenryG: 3) Is the VXLAN overlay network this MD supports described in the neutron network as a separate network segment from the VLAN segment that the openvswitch-agent MD binds with? | 13:47 |
| HenryG | rkukura: the vxlan overlay mentioned in the spec is outside of neutron | 13:48 |
| *** matsuhashi has joined #openstack-neutron | 13:50 | |
| rkukura | HenryG: OK, so then no new TypeDriver is needed. Does the NetworkEPG model map from network UUID to the overlay info? | 13:50 |
| HenryG | rkukura: correct, and yes | 13:51 |
| HenryG | rkukura: of course, in juno we plan to implement more advanced types, and Arvind has been in contact with you about that. | 13:53 |
| *** WackoRobie has joined #openstack-neutron | 13:53 | |
| rkukura | HenryG: Will this driver support multiple physical_networks for the VLAN links from ToR switch to nodes, or a single physical_network for this purpose? | 13:55 |
| *** WackoRobie has quit IRC | 13:55 | |
| *** thuc has joined #openstack-neutron | 13:56 | |
| *** thuc_ has joined #openstack-neutron | 13:56 | |
| rkukura | HenryG: And will which physical_network(s) it uses be configurable, so others can cooexist and be ignored by this MD? | 13:56 |
| rkukura | HenryG: Last question - are any enhancements to the ML2 plugin needed? | 13:57 |
| *** matsuhashi has quit IRC | 13:57 | |
| HenryG | rkukura: I think multiple physical links per node is supported, but I need to double check on that with the current code. | 13:58 |
| HenryG | rkukura: The physical networks are configured much like the cisco nexus MD. | 13:59 |
| HenryG | rkukura: no enhancments to ML2 plugin needed in icehouse. | 13:59 |
| rkukura | HenryG: Sounds good | 13:59 |
| *** thuc has quit IRC | 14:00 | |
| rkukura | HenryG: All set, but do you mind changing the title from "Cisco APIC Plugin for OpenStack Neutron" to "Cisco APIC Driver for OpenStack Neutron" to avoid confusion? | 14:02 |
| HenryG | rkukura: oops. Sure. (old habits) | 14:03 |
| *** matsuhashi has joined #openstack-neutron | 14:04 | |
| *** jecarey has joined #openstack-neutron | 14:04 | |
| mestery | enikanorov: I'll merge that one in if it's not hitting the crash, is it still so? | 14:09 |
| *** yfried has quit IRC | 14:11 | |
| jlibosva | hello, does anybody know if neutron-metadata-agent should report its hearbeats to neutron-server? from what I see in the code it should. From what I see in database, it doesn't. | 14:11 |
| *** Mierdin has quit IRC | 14:12 | |
| *** anand has quit IRC | 14:19 | |
| pcm | mestery: ping | 14:19 |
| mestery | pcm: pong | 14:21 |
| pcm | For a vendor driver I'm doing for VPNaaS, the device has a different ID requirements for the resources | 14:21 |
| pcm | mestery: So, I need to map a UUID from OS to and ID the device needs. | 14:22 |
| pcm | mestery: Was looking for ideas on how I can generate unique IDs, within a specific range, so that I can persist them in SQLAlchemy. | 14:22 |
| pcm | mestery: Any idea of who I could ask about that? (or any one in the list know) | 14:23 |
| pcm | mestery: For example, I may need to take UUIDs and translate them into the numbers 1..1000. | 14:23 |
| pcm | mestery: I know you can create an auto_increment field in SQL and make it unique, but don't know if one can have that wrap, or if that is even an appropriate way to do this. | 14:25 |
| mestery | pcm: So you just need to generate some UUIDs and store them as a mapping to an external ID? | 14:25 |
| pcm | mestery: I'll get a UUID and I need to generate an ID that will be compatible with the device. | 14:25 |
| mestery | pcm: I would ping a DB expert, the autoincrement field in SQL sounds like what you want | 14:26 |
| mestery | Especially if you can make it wrap | 14:26 |
| pcm | mestery: For example, OS may have an IKE policy with UUID "X" and I need to convert it to "1". The next UUID to "2". | 14:26 |
| pcm | mestery: Just don't know if that field can be set to wrap at a maximum (as the device has a limit on max value). | 14:27 |
| pcm | mestery: Know any OpenStack SQL experts? | 14:27 |
| mestery | pcm: enikanorov :) | 14:28 |
| *** matsuhashi has quit IRC | 14:28 | |
| pcm | enikanorov: ping :) | 14:28 |
| *** mrsnivvel has quit IRC | 14:29 | |
| *** changbl has joined #openstack-neutron | 14:31 | |
| openstackgerrit | Rossella Sblendido proposed a change to openstack/neutron: Introduce bulk calls for get device details https://review.openstack.org/66899 | 14:34 |
| *** yongli has quit IRC | 14:36 | |
| *** peristeri has joined #openstack-neutron | 14:37 | |
| *** Jabadia has joined #openstack-neutron | 14:40 | |
| *** thuc_ has quit IRC | 14:41 | |
| *** thuc has joined #openstack-neutron | 14:42 | |
| *** Jabadia has quit IRC | 14:44 | |
| *** thuc_ has joined #openstack-neutron | 14:45 | |
| *** thuc has quit IRC | 14:46 | |
| *** yamahata has joined #openstack-neutron | 14:50 | |
| *** yongli has joined #openstack-neutron | 14:52 | |
| *** rossella_s has quit IRC | 14:53 | |
| *** thuc_ has quit IRC | 14:53 | |
| *** thuc has joined #openstack-neutron | 14:54 | |
| *** thuc has quit IRC | 14:58 | |
| *** WackoRobie has joined #openstack-neutron | 15:00 | |
| *** russellb is now known as rustlebee | 15:02 | |
| *** Jabadia has joined #openstack-neutron | 15:08 | |
| *** pradipta has quit IRC | 15:09 | |
| *** pradipta has joined #openstack-neutron | 15:10 | |
| *** terryw has joined #openstack-neutron | 15:11 | |
| *** Jabadia has quit IRC | 15:13 | |
| *** rwsu has joined #openstack-neutron | 15:13 | |
| *** rossella_s has joined #openstack-neutron | 15:23 | |
| *** armax has joined #openstack-neutron | 15:26 | |
| openstackgerrit | Marcos Fermín Lobo proposed a change to openstack/python-neutronclient: Unexpected response in agent-list command https://review.openstack.org/70363 | 15:28 |
| *** mflobo has quit IRC | 15:29 | |
| anteaya | ygbo: well taht patch hit https://bugs.launchpad.net/nova/+bug/1254890 | 15:32 |
| anteaya | so recheck bug 1254890 is the preferred syntax | 15:32 |
| anteaya | and yes, rechecking is fine | 15:33 |
| anteaya | thanks for asking | 15:33 |
| *** markmcclain has joined #openstack-neutron | 15:33 | |
| ygbo | anteaya: thanks :-) | 15:36 |
| anteaya | np | 15:36 |
| *** IanGovett1 has quit IRC | 15:38 | |
| *** MM_at_HP has joined #openstack-neutron | 15:38 | |
| *** IanGovett has joined #openstack-neutron | 15:39 | |
| *** mfink has joined #openstack-neutron | 15:39 | |
| *** AlexF_ has joined #openstack-neutron | 15:43 | |
| *** morganfainberg|z is now known as morganfainberg | 15:43 | |
| *** unicell has quit IRC | 15:45 | |
| *** unicell1 has quit IRC | 15:45 | |
| *** unicell has joined #openstack-neutron | 15:45 | |
| *** jpich has quit IRC | 15:47 | |
| *** unicell_ has joined #openstack-neutron | 15:48 | |
| *** thedodd has joined #openstack-neutron | 15:48 | |
| *** unicell_ has quit IRC | 15:50 | |
| haleyb | jlibosva: hi jakub. i saw you reviewed my change, thanks, i'll get an update out | 15:51 |
| *** alagalah has joined #openstack-neutron | 15:52 | |
| *** alex_klimov has quit IRC | 15:52 | |
| *** pasquier-s has quit IRC | 15:55 | |
| *** emagana has joined #openstack-neutron | 15:56 | |
| openstackgerrit | Sean M. Collins proposed a change to openstack/neutron: Create new IPv6 attributes for Subnets https://review.openstack.org/52983 | 15:56 |
| sc68cal | this time with the correct down_revision for the alembic script :-\ | 15:58 |
| *** markwash has joined #openstack-neutron | 15:59 | |
| *** emagana has quit IRC | 16:01 | |
| *** beagles is now known as seagulls | 16:01 | |
| *** doddstack has joined #openstack-neutron | 16:02 | |
| *** salv-orlando has joined #openstack-neutron | 16:03 | |
| *** vkozhukalov has quit IRC | 16:04 | |
| jlibosva | haleyb: Hi, I'm working on optimization of queries to neutron. I created a cache for requests with small expiration time. I don't have unittests yet and I didn't test it properly. I'll push the patch based on yours once it's finished. | 16:04 |
| *** thedodd has quit IRC | 16:05 | |
| markmcclain | jlibosva: how long are you caching? | 16:08 |
| jlibosva | markmcclain: configurable, I was thinking about 5 seconds | 16:08 |
| jlibosva | by default | 16:08 |
| haleyb | jlibosva: ok, thanks, i'll review it. just updating mine now | 16:09 |
| markmcclain | ok.. the current allocator is likely to reuse an address fairly quickly, so the longer we make it higher the chance of it getting out of sync | 16:09 |
| openstackgerrit | Sean M. Collins proposed a change to openstack/neutron: Create new IPv6 attributes for Subnets https://review.openstack.org/52983 | 16:09 |
| *** sweston has quit IRC | 16:09 | |
| jlibosva | I'm not very familiar with cloud-init. What I saw is that it passed router-id and instance's ip-address. I don't understand where does this router id come from. | 16:11 |
| *** sandeepr has left #openstack-neutron | 16:16 | |
| *** aymenfrikha has quit IRC | 16:18 | |
| openstackgerrit | Brian Haley proposed a change to openstack/neutron: Change metadata-agent to spawn multiple workers https://review.openstack.org/70204 | 16:20 |
| *** mlavalle has joined #openstack-neutron | 16:21 | |
| *** aymenfrikha has joined #openstack-neutron | 16:26 | |
| *** SumitNaiksatam has quit IRC | 16:28 | |
| markmcclain | jlibosva: the router_id is from the metadata proxy that's gets the info from its command line args | 16:29 |
| *** jistr has quit IRC | 16:30 | |
| jlibosva | markmcclain: I don't understand this. I thought cloud-init calls the metadata proxy and it provides the router-id in the header of this call. | 16:30 |
| markmcclain | no cloud init provides no data in it's header | 16:31 |
| jlibosva | markmcclain: who is the caller of metadata proxy then? | 16:32 |
| markmcclain | the instance is the caller of the namespace metadata proxy | 16:32 |
| markmcclain | it makes a plain HTTP call to 169.254.169.254 and we intercept that | 16:33 |
| markmcclain | the namespace metadata proxy then adds headers | 16:33 |
| markmcclain | https://git.openstack.org/cgit/openstack/neutron/tree/neutron/agent/metadata/namespace_proxy.py#n81 | 16:33 |
| markmcclain | and passes the request via the UnixDomain socket outside of the namespace to the Neutron metadata agent which forwards the request to Nova | 16:33 |
| openstackgerrit | Marios Andreou proposed a change to openstack/neutron: Make allocation_pools attribute of subnet updateable by PUT https://review.openstack.org/62042 | 16:34 |
| openstackgerrit | Brian Haley proposed a change to openstack/neutron: Change metadata-agent to have a configurable backlog https://review.openstack.org/70214 | 16:35 |
| jlibosva | markmcclain: ahaa, thanks for explanation :) | 16:38 |
| markmcclain | jlibosva: this workflow something we need to doc a little a better | 16:42 |
| *** aymenfrikha has quit IRC | 16:43 | |
| jlibosva | yeah, I still struggle with workflows :) | 16:43 |
| *** aymenfrikha has joined #openstack-neutron | 16:48 | |
| *** banix has joined #openstack-neutron | 16:53 | |
| *** aburaschi1 has joined #openstack-neutron | 16:54 | |
| *** jprovazn has quit IRC | 16:55 | |
| *** aburaschi has quit IRC | 16:55 | |
| openstackgerrit | Rossella Sblendido proposed a change to openstack/neutron: Introduce bulk calls for get device details https://review.openstack.org/66899 | 16:58 |
| *** safchain has quit IRC | 17:00 | |
| *** markmcclain has quit IRC | 17:01 | |
| *** sbalukoff has quit IRC | 17:02 | |
| *** snowblind2 has joined #openstack-neutron | 17:03 | |
| *** networkstatic has joined #openstack-neutron | 17:04 | |
| banix | Hi everybody; Have a patch that fails gate-neutron-python26 and 27. Looking at the console output for those tests I do not see any failure related to my patch. Here is what I see: FAIL: neutron.tests.unit.test_provider_configuration.ParseServiceProviderConfigurationTestCase.test_parse_service_provider_opt_not_allowed_raises | 17:05 |
| banix | I thought I saw a bug related to this on the rechecks site but I cannot find anything related on that page anymore. So I am wondering if this is an issue that others have as well or I am missing something and my patch is the source of the failure | 17:06 |
| *** snowblind2 has quit IRC | 17:08 | |
| *** snowblind2 has joined #openstack-neutron | 17:08 | |
| *** ygbo has quit IRC | 17:08 | |
| *** markmcclain has joined #openstack-neutron | 17:09 | |
| *** sweston has joined #openstack-neutron | 17:11 | |
| marun | banix: there is a patch pending merge that causes unit test failures: https://review.openstack.org/#/c/67537/ | 17:12 |
| sc68cal | I'm hitting the same unit test failure - glad it's not something I broke | 17:13 |
| banix | Thanks for responding; This has been ongoing for a couple of weeks or so. So I thought may be I am missing something and it is my patch. | 17:15 |
| banix | Thanks for working on the patch. It looks like it is ready to be merged. | 17:15 |
| asadoughi | marun: so, it's just waiting for a +A? | 17:15 |
| marun | asadoughi: yes. | 17:16 |
| marun | markmcclain: is there a timeline on starting to approve critical fixes? | 17:16 |
| markmcclain | marun: we've got 2 patches (1 Nova and 1 Devstack to get merged first) | 17:17 |
| *** changbl has quit IRC | 17:17 | |
| markmcclain | once those merge, I'll work on getting the prioritized reviews merged | 17:17 |
| marun | markmcclain: thanks for the update | 17:17 |
| markmcclain | depending on when those get merged, I'll coordinate with the cores who are online here in the channel | 17:18 |
| banix | marun: Thanks. | 17:18 |
| marun | markmcclain: ok | 17:18 |
| markmcclain | the translation patch is currently #1 on my list | 17:18 |
| *** emagana has joined #openstack-neutron | 17:20 | |
| *** markwash has quit IRC | 17:23 | |
| seagulls | salv-orlando, sanity check - port updates in ovs_neutron_agent are only done within the execution context of the agent's rpc_loop, correct? | 17:25 |
| seagulls | (or will be .. pending one of your patches) | 17:26 |
| seagulls | salv-orlando, oh forget it | 17:26 |
| salv-orlando | ok | 17:26 |
| openstackgerrit | Brian Haley proposed a change to openstack/neutron: Change metadata-agent to spawn multiple workers https://review.openstack.org/70204 | 17:26 |
| seagulls | salv-orlando, just saw the "Depends On" in the review :) | 17:26 |
| salv-orlando | ok, no worries | 17:26 |
| seagulls | salv-orlando, kind of obvious now .. heh | 17:26 |
| *** sbalukoff has joined #openstack-neutron | 17:31 | |
| *** mfink has quit IRC | 17:32 | |
| *** sweston has quit IRC | 17:32 | |
| *** terryw is now known as otherwisegu | 17:34 | |
| *** otherwisegu is now known as otherwiseguy | 17:34 | |
| *** sweston has joined #openstack-neutron | 17:35 | |
| *** rohit404 has quit IRC | 17:37 | |
| *** emagana has quit IRC | 17:38 | |
| *** markwash has joined #openstack-neutron | 17:39 | |
| *** bjornar has joined #openstack-neutron | 17:40 | |
| sc68cal | banix: I think I see the bug in the rechecks page now | 17:44 |
| sc68cal | bug 1270212 | 17:44 |
| *** Mierdin has joined #openstack-neutron | 17:45 | |
| *** Guest6335 has quit IRC | 17:46 | |
| banix | sc68cal: Thanks. Yes that's the one; I remember the bug number :) | 17:46 |
| sc68cal | ok - did you add a recheck to your review to link to that bug? | 17:47 |
| banix | I had done it earlier on; This has been going on for a couple of weeks. | 17:47 |
| sc68cal | ah ok | 17:47 |
| marun | pcm: ping | 17:47 |
| *** simon-AS5591 has quit IRC | 17:48 | |
| *** simon-AS559 has joined #openstack-neutron | 17:48 | |
| *** baoli has quit IRC | 17:51 | |
| *** simon-AS559 has quit IRC | 17:52 | |
| *** carl_baldwin has joined #openstack-neutron | 17:54 | |
| *** harlowja_away is now known as harlowja | 17:55 | |
| *** markwash has quit IRC | 17:56 | |
| *** bjornar has quit IRC | 17:57 | |
| *** baoli has joined #openstack-neutron | 17:58 | |
| *** shivh has joined #openstack-neutron | 18:00 | |
| *** SumitNaiksatam has joined #openstack-neutron | 18:03 | |
| *** emagana has joined #openstack-neutron | 18:05 | |
| *** sdague_ has joined #openstack-neutron | 18:10 | |
| *** sdague has quit IRC | 18:10 | |
| *** sdague_ is now known as sdague | 18:11 | |
| *** carl_baldwin has quit IRC | 18:11 | |
| *** godara has joined #openstack-neutron | 18:12 | |
| *** markmcclain has quit IRC | 18:12 | |
| *** rossella_s has quit IRC | 18:14 | |
| *** shashank_ has joined #openstack-neutron | 18:15 | |
| *** dims has quit IRC | 18:15 | |
| *** carl_baldwin has joined #openstack-neutron | 18:17 | |
| *** dims has joined #openstack-neutron | 18:18 | |
| *** baoli has quit IRC | 18:18 | |
| *** doddstack has quit IRC | 18:20 | |
| *** yamahata has quit IRC | 18:21 | |
| *** jlibosva has quit IRC | 18:22 | |
| anteaya | I'm airporting today - marun and markmcclain are you best people to discuss rechecks and the status of the gate | 18:24 |
| anteaya | thanks marun | 18:25 |
| *** sn6i23a has joined #openstack-neutron | 18:25 | |
| *** humbolt has quit IRC | 18:26 | |
| *** sweston has quit IRC | 18:28 | |
| anteaya | and thanks sc68cal for jumping on the elastic-recheck train | 18:29 |
| anteaya | welcome aboard | 18:29 |
| *** shivh has quit IRC | 18:30 | |
| *** AlexF_ has quit IRC | 18:30 | |
| *** Jabadia has joined #openstack-neutron | 18:31 | |
| *** jog0 is now known as flashgordon | 18:32 | |
| *** Jabadia has joined #openstack-neutron | 18:32 | |
| *** sweston has joined #openstack-neutron | 18:35 | |
| *** bjornar has joined #openstack-neutron | 18:36 | |
| *** aburaschi1 is now known as aburaschi | 18:38 | |
| enikanorov__ | mestery: ping | 18:39 |
| *** shivh has joined #openstack-neutron | 18:41 | |
| *** carl_baldwin has quit IRC | 18:42 | |
| marun | anteaya: will do my best | 18:43 |
| openstackgerrit | Sean M. Collins proposed a change to openstack/neutron: Create new IPv6 attributes for Subnets https://review.openstack.org/52983 | 18:44 |
| * sc68cal grumbles | 18:46 | |
| sc68cal | I'm being stupid - missing pep8 checks. | 18:46 |
| marun | sc68cal: happens to the best of us | 18:48 |
| sc68cal | anteaya: thanks for the welcome | 18:49 |
| *** baoli has joined #openstack-neutron | 18:49 | |
| *** djoreilly has quit IRC | 18:49 | |
| marun | sc68cal: that kind of mechanical stuff should be automatic, frankly. | 18:49 |
| sc68cal | I use python-mode for vim, which has pep8 included. Although i think something died in my vim session and it stopped rechecking | 18:49 |
| sc68cal | since it didn't complain about the indentation | 18:50 |
| marun | sc68cal: well, pep8 does all kinds of hacking checks, too | 18:50 |
| marun | sc68cal: or is the mode running flake8? | 18:50 |
| sc68cal | yeah flake8 - sorry | 18:51 |
| marun | ah, ok | 18:51 |
| sc68cal | although I don't think it picks up the config that we have for openstack projects | 18:51 |
| marun | i suppose i should enable that | 18:51 |
| sc68cal | marun: you should - and also check out YouCompleteMe | 18:51 |
| sc68cal | makes your vim look like this - https://pbs.twimg.com/media/BavaGdwCcAAE5O9.png:large | 18:52 |
| marun | sc68cal: yeah, I don't like autocompletion | 18:52 |
| marun | sc68cal: The thing I find most useful is jump-to-definition, via rope | 18:53 |
| *** jgrimm has joined #openstack-neutron | 18:53 | |
| sc68cal | I had problems with rope (https://github.com/klen/python-mode/issues/342) - so I use ctags. | 18:54 |
| marun | sc68cal: ah, that sucks | 18:55 |
| marun | ctags is a poor substitute | 18:55 |
| *** seagulls is now known as beagles | 18:56 | |
| *** baoli has quit IRC | 18:57 | |
| *** WackoRobie has quit IRC | 18:58 | |
| *** jaypipes has quit IRC | 18:59 | |
| *** godara has quit IRC | 18:59 | |
| *** beagles is now known as seagulls | 19:00 | |
| *** dims has quit IRC | 19:00 | |
| *** godara has joined #openstack-neutron | 19:01 | |
| *** AlexF_ has joined #openstack-neutron | 19:02 | |
| *** networkstatic has quit IRC | 19:02 | |
| *** markwash has joined #openstack-neutron | 19:02 | |
| *** sweston has quit IRC | 19:05 | |
| *** thedodd has joined #openstack-neutron | 19:06 | |
| *** markwash has quit IRC | 19:08 | |
| *** jaypipes has joined #openstack-neutron | 19:09 | |
| *** dims has joined #openstack-neutron | 19:14 | |
| *** markwash has joined #openstack-neutron | 19:15 | |
| *** thuc has joined #openstack-neutron | 19:18 | |
| *** WackoRobie has joined #openstack-neutron | 19:18 | |
| *** vkozhukalov has joined #openstack-neutron | 19:19 | |
| *** thuc has quit IRC | 19:20 | |
| *** markmcclain has joined #openstack-neutron | 19:20 | |
| *** thuc has joined #openstack-neutron | 19:20 | |
| *** pradipta has quit IRC | 19:22 | |
| *** thuc has quit IRC | 19:25 | |
| *** AlexF_ has quit IRC | 19:29 | |
| *** baoli has joined #openstack-neutron | 19:33 | |
| sc68cal | emagana: Hey, I registered a BP for docs to make some more developer docs for Neutron - https://blueprints.launchpad.net/neutron/+spec/developer-documentation | 19:34 |
| sc68cal | I've got some people on the Comcast team that I'm onboarding into the Neutron codebase and I'm probably going to do a brain dump of all the parts of neutron I understand, so they can get up to speed quicker than i did | 19:35 |
| emagana | sc68cal: sounds great! | 19:43 |
| emagana | sc68cal: if you want a 1:1 session to understand anything just let me know | 19:43 |
| sc68cal | OK - will do. I'm going to focus first on the api extensions part of Neutron since that is where I cut my teeth. | 19:44 |
| sc68cal | I'll probably lean heavily on salv-orlando's presentation about writing a new plugin that he presented at HK | 19:44 |
| sc68cal | just dig up the link since that seemed to be pretty comprehensive | 19:44 |
| *** nati_ueno has joined #openstack-neutron | 19:50 | |
| *** nati_ueno has quit IRC | 19:50 | |
| *** sweston has joined #openstack-neutron | 19:51 | |
| *** nati_ueno has joined #openstack-neutron | 19:51 | |
| *** changbl has joined #openstack-neutron | 19:56 | |
| sc68cal | mestery: Do you have any dev doc for the ml2 plugin internals? Worst case I may just scrape your slideshare URLs for that | 19:56 |
| dkehn | enikanorov__: u around, looking at https://review.openstack.org/#/c/55032/10 | 19:57 |
| mestery | sc68cal: We did at one point, let me see if I can dig it out, though it was somewhat highlevel I think. | 19:57 |
| sc68cal | mestery: that's OK - better something than nothing, and I'd prefer high level, since we can drill down later | 19:57 |
| mestery | :) | 19:57 |
| enikanorov__ | dkehn: ok | 19:58 |
| enikanorov__ | mestery: Hi | 19:58 |
| *** sweston has quit IRC | 19:58 | |
| mestery | enikanorov__: Hey, sorry for missing each other, today has been as crazy as yesterday on many fronts. :) | 19:58 |
| mestery | Whats up? | 19:58 |
| *** otherwiseguy has quit IRC | 19:58 | |
| enikanorov__ | mestery: no prob. i thought it might make sense to approve https://review.openstack.org/#/c/66670/ since it seem passing checks consistently | 19:59 |
| mestery | enikanorov__: Yes, I agree, from reading the scrollback, looks like markmcclain was saying we should prioritize things like this. | 19:59 |
| mestery | markmcclain: Are we good to approve https://review.openstack.org/#/c/66670/? | 19:59 |
| dkehn | enikanorov__: https://review.openstack.org/#/c/55032/10/neutron/tests/unit/services/loadbalancer/drivers/haproxy/test_plugin_driver.py#L636 I'm getting a TypeError becuase the p p['members']= [u'7c52b7aa-ac97-4df1-8b26-563d68b2d42d'] | 20:00 |
| enikanorov__ | dkehn: yeah, for some reason I missed UT failures... looks like I've looked at py27 and 26 failures and thought that it is that issue with provider configuration/exceptions | 20:02 |
| enikanorov__ | dkehn: I'll fix the patch | 20:02 |
| dkehn | enikanorov__: thx | 20:02 |
| enikanorov__ | thanks for letting me know | 20:03 |
| *** emagana has quit IRC | 20:03 | |
| markmcclain | mestery: not yet..we've got a dependent devstack change to merge first | 20:04 |
| mestery | markmcclain: Thanks for the confirmation! | 20:04 |
| markmcclain | mestery: this is also on my list of reviews to approve once we're ready to open everything back up | 20:06 |
| mestery | Cool, thanks markmcclain! | 20:06 |
| *** changbl has quit IRC | 20:07 | |
| *** changbl has joined #openstack-neutron | 20:07 | |
| *** sweston has joined #openstack-neutron | 20:08 | |
| *** Jabadia has quit IRC | 20:10 | |
| *** godara has quit IRC | 20:12 | |
| *** sweston has quit IRC | 20:16 | |
| enikanorov__ | markmcclain: what is the devstack change the neutron depends on? | 20:16 |
| enikanorov__ | btw, was kernel crash issue resolved? | 20:17 |
| enikanorov__ | if so, how? | 20:17 |
| mestery | enikanorov__: Yes, checkout out the launchpad bug regarding kernel crash. | 20:18 |
| enikanorov__ | ok | 20:18 |
| mestery | Ended up being NMB volume being mounted on the guest at the same time as network namespace commands are being issued. | 20:18 |
| mestery | salv-orlando is the rockstar who found that guy finally :) | 20:19 |
| mestery | salv and I were running scripts all afternoon yesterday to try and reproduce until he hit on the trigger | 20:19 |
| mestery | At one point, we each had over a thousand namespaces and dnsmasqs running :) | 20:19 |
| *** godara has joined #openstack-neutron | 20:20 | |
| pcm | enikanorov__: ping | 20:22 |
| enikanorov__ | mestery: yep, i've read the logs :) | 20:23 |
| enikanorov__ | pcm: pong | 20:23 |
| pcm | enikanorov__: Kyle suggested I contact you for some questions on database. Have a few mins? | 20:23 |
| enikanorov__ | mestery: i'm wondering why check jobs started to pass | 20:23 |
| enikanorov__ | pcm: sure | 20:23 |
| * pcm sorry Kyle :) | 20:23 | |
| *** otherwiseguy has joined #openstack-neutron | 20:24 | |
| pcm | enikanorov__: I have to convert UUIDs to another ID that has a limited (integer) range for a vendor plugin on VPNaaS | 20:24 |
| pcm | enikanorov__: Was wondering how to generate the ID and keep it within range. | 20:24 |
| pcm | enikanorov__: Could I use database and auto_increment? | 20:25 |
| pcm | enikanorov__: (I have to persist the values for later delete operations) | 20:25 |
| enikanorov__ | well, if the range is not very small, i think it's better to get a portion of UUID | 20:25 |
| enikanorov__ | that fit into that ID range | 20:25 |
| pcm | enikanorov__: One can take a 31 char string, others have to be integers and one needs to be 1..N (and I think N is small) | 20:26 |
| enikanorov__ | and you either rely on uniqueness or add some logic to enforce it | 20:26 |
| enikanorov__ | oh, it's integer... | 20:26 |
| pcm | enikanorov__: two are, the other can be a string, just not 36 chars long :( | 20:26 |
| *** rohit404 has joined #openstack-neutron | 20:27 | |
| enikanorov__ | 31 chars has enough space for universe uniqueness, so it will do. for integer - yeah, auto_increment should work | 20:27 |
| pcm | enikanorov__: Will it wrap though, at some value? | 20:28 |
| pcm | enikanorov__: I'm checking to find out max int value, but let's say it's 16 bit or less | 20:28 |
| pcm | enikanorov__: If I create a database field and use auto_increment, can I enforce that it wraps? | 20:29 |
| enikanorov__ | i think you need to set field to a proper type | 20:29 |
| pcm | enikanorov__: so select some type within the limits and then use auth_increment? | 20:30 |
| enikanorov__ | do you know the limits? | 20:30 |
| pcm | enikanorov__: Not yet, trying to find out. | 20:31 |
| * pcm pinging failed...trying email | 20:34 | |
| *** vkozhukalov has quit IRC | 20:34 | |
| *** sweston has joined #openstack-neutron | 20:35 | |
| *** Jabadia has joined #openstack-neutron | 20:36 | |
| pcm | enikanorov__: I have the ranges... | 20:39 |
| enikanorov__ | what is the range? | 20:39 |
| pcm | enikanorov__: One is 31 character, one is 1..10,000, and one is 1..7FFFFFF | 20:39 |
| enikanorov__ | ok, so the last one is regular Integer | 20:40 |
| enikanorov__ | so auto_increment should work | 20:40 |
| *** Jabadia has quit IRC | 20:40 | |
| pcm | enikanorov__: cool | 20:40 |
| pcm | enikanorov__: Can I chop the UUID for the first one (the first or last 31?)? | 20:41 |
| *** Jabadia has joined #openstack-neutron | 20:41 | |
| enikanorov__ | as for 1..10000 - auto_increment will work as well, but you'll need to check if you're within the range | 20:41 |
| enikanorov__ | eys, i think chopping will work, 31 chars is enough to be safe | 20:42 |
| *** baoli has quit IRC | 20:42 | |
| pcm | enikanorov__: Chop the first 31 chars OK? | 20:42 |
| enikanorov__ | yes | 20:42 |
| pcm | enikanorov__: For the 1..10000, so I tell the DBase to use auto_increment, but how do I make sure it wraps at 10000? | 20:43 |
| enikanorov__ | check this one: http://stackoverflow.com/questions/10494033/setting-sqlalchemy-autoincrement-start-value | 20:43 |
| enikanorov__ | not extactly what you're looking for | 20:43 |
| enikanorov__ | but i guess end of the range should be declared somehow similar | 20:44 |
| pcm | enikanorov__: So you think I check the value and if >10000, set it to 1? Wondering how I ensure unique though. | 20:46 |
| *** Jabadia has quit IRC | 20:46 | |
| enikanorov__ | pcm: there's no simple way to ensure uniqueness. auto_incremet assumes that you have well enough space for all ids you need | 20:46 |
| *** ajo has quit IRC | 20:47 | |
| enikanorov__ | another way is to do similar to ipallocation logic | 20:47 |
| enikanorov__ | availability ranges, etc | 20:47 |
| enikanorov__ | e.g. pretty complex | 20:47 |
| pcm | enikanorov__: In reality, there will probably only be a few of these. They are for the IKE policy, and I can't imaging a tenant having > 10. | 20:48 |
| enikanorov__ | auto_increment is table-wise | 20:48 |
| pcm | enikanorov__: could I use auto_increment with a small type, like a byte? | 20:48 |
| enikanorov__ | pcm: yes | 20:49 |
| pcm | enikanorov__: Maybe I can do that and just fail the request if they have > 256 | 20:49 |
| pcm | 255 | 20:49 |
| enikanorov__ | but that will not work. one tenant has < 10 ikepolicies, 26 tenants will have 260 | 20:50 |
| pcm | enikanorov__: Will auto_increment fill in the holes and use the next available value, or will it wrap and duplicate? | 20:50 |
| enikanorov__ | no, it will not fill the holes, i guess it will generate exception | 20:52 |
| enikanorov__ | instead of wrapping | 20:52 |
| *** rohit404 has quit IRC | 20:52 | |
| pcm | enikanorov__: And it won't ensure uniqueness? | 20:55 |
| enikanorov__ | it just goes through the range. so it kinda ensures it, but in your case of small range it's not very good way of doing it | 20:56 |
| pcm | enikanorov__: Ay other ideas on how I should ensure uniqueness with a small range of values? | 20:57 |
| enikanorov__ | yeah | 20:57 |
| pcm | enikanorov__: Should I somehow augment with python code? | 20:57 |
| enikanorov__ | i don't know whether it a common practice | 20:57 |
| enikanorov__ | so if you have 1..10000 range only | 20:58 |
| pcm | enikanorov__: Like, have a transaction, and then get all values and determine the next to use. | 20:58 |
| enikanorov__ | you may store bit map | 20:58 |
| enikanorov__ | of 10000 bits | 20:58 |
| pcm | interesting. | 20:58 |
| enikanorov__ | it's a tradeoff between code complexity vs performance. finding a hole may take up to N cycles (where N=10000) | 20:59 |
| enikanorov__ | smarter approach will be more complex | 20:59 |
| enikanorov__ | but with small ranges it may make sense | 20:59 |
| pcm | so have this map persisted and look for a hole and use that for the ID? | 21:00 |
| enikanorov__ | if bit M = 0, then M is free ID | 21:01 |
| enikanorov__ | and you mark it | 21:01 |
| *** ajo has joined #openstack-neutron | 21:01 | |
| pcm | enikanorov__: store that map in another table? | 21:01 |
| enikanorov__ | not really. store it in memory and sync with the table on startup | 21:02 |
| pcm | enikanorov__: or you think build it at startup from datbase and keep in mem? | 21:02 |
| enikanorov__ | sure | 21:02 |
| pcm | Even the Integer field (0.7FFFFFFF) won't have a ton of entries. It represents the IPSec connections, so can't imagine a ton of those. | 21:03 |
| sc68cal | nmb? | 21:05 |
| *** baoli has joined #openstack-neutron | 21:05 | |
| sc68cal | whoops - sorry, | 21:05 |
| *** aburaschi has left #openstack-neutron | 21:06 | |
| *** baoli has quit IRC | 21:06 | |
| *** baoli has joined #openstack-neutron | 21:07 | |
| enikanorov__ | pcm: for 0..7ffffff that simple approach will not work | 21:16 |
| pcm | enikanorov__: the auto_increment? | 21:16 |
| enikanorov__ | no, i mean bit map | 21:17 |
| enikanorov__ | auto_increment there should be fine | 21:17 |
| pcm | enikanorov__: I'm just wondering if I should use a single bit map, limit it to say 1000 and use that for all three IDs. | 21:18 |
| enikanorov__ | not sure i get the idea | 21:19 |
| pcm | enikanorov__: We won't have 1000 IKE policies, IPSec policies, and IPSec tunnels for a tenant. | 21:19 |
| pcm | Those IDs represent those three values. Real world, probably only a few of the policies, and probably under 10 tunnels for one tenant (site to site tunnels) | 21:20 |
| pcm | so, although the device has limited ranges, we'll never see those many practically. | 21:21 |
| enikanorov__ | bit map is only suitable for tunnel ids, because it should span whole tunnel id range | 21:21 |
| enikanorov__ | and the only reason to use bit map is to simplify code. smaller range the better | 21:21 |
| enikanorov__ | for other ids auto_increment or a part of UUID should work fine | 21:22 |
| pcm | The tunnel is the one that is 0..7FFFFFFF. The IKE policy is limited to 10000, the IPSec policy is limited to 31 char | 21:22 |
| pcm | The same ID can be used for each, as long as they are unique among others IDs of the same type. | 21:23 |
| *** sweston has quit IRC | 21:24 | |
| pcm | Thinking... I could use a bit array, set it to 1000 (or 2000, or 10000). | 21:24 |
| pcm | use that for the tunnel. I guess since the policies are not created at the same time, use a smaller bit array for the IKE policy, and use the UUID trimmed for IPSec. | 21:25 |
| pcm | enikanorov__: great idea about the bit array. | 21:26 |
| enikanorov__ | well, it's not that i just invented it :) | 21:28 |
| pcm | enikanorov__: Make sense (did I explain it clearly)? | 21:28 |
| pcm | enikanorov__: Yeah, but I had blinders on, thinking of database only solution. | 21:28 |
| openstackgerrit | rcurran proposed a change to openstack/neutron: ML2 Cisco Nexus MD: Create pre/post DB event handlers https://review.openstack.org/56478 | 21:29 |
| pcm | enikanorov__: Another question about the dbase... | 21:29 |
| pcm | enikanorov__: For the vendor driver, I need to persist these three IDs. Do I just create a new table from the service driver? | 21:29 |
| enikanorov__ | from the service driver? what do you mean 'from' ? | 21:30 |
| *** sweston has joined #openstack-neutron | 21:30 | |
| pcm | enikanorov__: Have the service driver create and populate the new table? | 21:30 |
| enikanorov__ | you mean the table definition is in driver's file? | 21:30 |
| * pcm wondering if it is as simple as that | 21:30 | |
| enikanorov__ | i think it's the same way is any other table usage | 21:31 |
| enikanorov__ | you need migration and definition | 21:31 |
| pcm | enikanorov__: Yeah, define the table in the service driver on the plugin. | 21:31 |
| pcm | enikanorov__: I don't try to extend the "reference" tables that exist? | 21:31 |
| enikanorov__ | the table you are creating is vendor-specific, right? if so, it gonna be separate, and created with a migration | 21:32 |
| pcm | enikanorov__: yes. ok. Just wanted to make sure there wasn't some other method | 21:33 |
| * pcm I couldn't think of another way | 21:33 | |
| enikanorov__ | me too | 21:33 |
| pcm | enikanorov__: If there is a failure on the create of vendor table entry, what happens on the reference table entry? | 21:34 |
| pcm | enikanorov__: IOW, how to I ensure they are consistent? | 21:34 |
| enikanorov__ | that not very clear to me | 21:34 |
| enikanorov__ | what is reference table? do you have a code that implements that? | 21:34 |
| pcm | the table that the VPN plugin has | 21:35 |
| pcm | (non vendor stuff) | 21:35 |
| pcm | The flow seems to be create the entry in the VPN plugin table, look up the provider's service driver, call it's create fucntion, which I guess is where I'd create the vendor table and then message the device driver (if applicable) | 21:36 |
| pcm | https://github.com/openstack/neutron/blob/master/neutron/services/vpn/plugin.py#L52 | 21:37 |
| pcm | https://github.com/openstack/neutron/blob/master/neutron/db/vpn/vpn_db.py#L230 | 21:38 |
| pcm | wondering, if the VPNPluginDB creates the ipsec_site_connections table, and then calls the vendor service driver, I'd create my table entry there. | 21:40 |
| pcm | at that point, if it fails, wondering how I roll back the ipsec_site_connections table entry? | 21:40 |
| *** Jabadia has joined #openstack-neutron | 21:42 | |
| openstackgerrit | Praneet Bachheti proposed a change to openstack/neutron: Juniper Contrail plug-in implementation for core resources https://review.openstack.org/43793 | 21:43 |
| *** Jabadia has quit IRC | 21:46 | |
| pcm | enikanorov__: Does that make sense? | 21:49 |
| *** ajo has quit IRC | 21:49 | |
| enikanorov__ | pcm: looking | 21:50 |
| pcm | enikanorov__: thanks. Not sure I was being clear over IRC | 21:50 |
| *** carl_baldwin has joined #openstack-neutron | 21:52 | |
| *** emagana has joined #openstack-neutron | 21:52 | |
| enikanorov__ | pcm: "which I guess is where I'd create the vendor table " | 21:52 |
| enikanorov__ | you mean an entry in vendor's table? | 21:52 |
| enikanorov__ | because table itself should be created at startup i guess / (or with migration) | 21:53 |
| pcm | enikanorov__: yes | 21:53 |
| *** openstackgerrit has quit IRC | 21:53 | |
| enikanorov__ | ok | 21:53 |
| *** armax has left #openstack-neutron | 21:53 | |
| *** openstackgerrit has joined #openstack-neutron | 21:53 | |
| enikanorov__ | that makes sense to me | 21:53 |
| pcm | enikanorov__: so if there is a failure there, the transaction on the ipsec_site_connections has already completed, right? | 21:53 |
| pcm | enikanorov__: so, i'd have one table with entry and one without? | 21:54 |
| openstackgerrit | Praneet Bachheti proposed a change to openstack/neutron: Juniper Contrail plug-in implementation for core resources https://review.openstack.org/43793 | 21:54 |
| enikanorov__ | usually if you fail to deploy logical config to backend (and deployment is reflected with vendor-specific table) than you put config in ERROR state | 21:55 |
| enikanorov__ | so in case it is in ERROR, such inconsistency between 'plugin table' and 'driver table' might be expected state | 21:55 |
| *** sweston has quit IRC | 21:56 | |
| pcm | "put config in ERROR state" == set the status field in the plugin table to mark that as being in error? | 21:57 |
| openstackgerrit | Praneet Bachheti proposed a change to openstack/neutron: Juniper Contrail plug-in implementation for core resources https://review.openstack.org/43793 | 22:02 |
| enikanorov__ | pcm: | 22:05 |
| enikanorov__ | pcm: yes | 22:05 |
| pcm | then, the user could delete and re-create the config (thereby fixing up the plugin table)? | 22:05 |
| enikanorov__ | i'm half asleep now.... | 22:06 |
| enikanorov__ | pcm. yes, or update it with admin_state_up to force redeploy | 22:06 |
| enikanorov__ | (if that's how fwaas works) | 22:06 |
| pcm | thanks for all your help!!!! | 22:07 |
| enikanorov__ | ok, i'm going to fall onto my pillow now | 22:07 |
| enikanorov__ | np, see you | 22:07 |
| openstackgerrit | Praneet Bachheti proposed a change to openstack/neutron: Juniper Contrail plug-in implementation for core resources https://review.openstack.org/43793 | 22:10 |
| *** pcm has quit IRC | 22:14 | |
| *** baoli has quit IRC | 22:19 | |
| *** Apsu is now known as SenorSnottyTits | 22:20 | |
| *** SenorSnottyTits is now known as Apsu | 22:21 | |
| *** carl_baldwin has quit IRC | 22:24 | |
| *** carl_baldwin has joined #openstack-neutron | 22:33 | |
| *** peristeri has quit IRC | 22:42 | |
| *** Jabadia has joined #openstack-neutron | 22:43 | |
| *** Jabadia has quit IRC | 22:45 | |
| *** Jabadia has joined #openstack-neutron | 22:45 | |
| *** jecarey has quit IRC | 22:46 | |
| *** carl_baldwin has quit IRC | 22:49 | |
| *** Jabadia has quit IRC | 22:50 | |
| *** carl_baldwin has joined #openstack-neutron | 22:59 | |
| *** mlavalle has quit IRC | 23:04 | |
| *** jgrimm has quit IRC | 23:06 | |
| *** carl_baldwin has quit IRC | 23:07 | |
| *** dims has quit IRC | 23:08 | |
| *** IanGovett has quit IRC | 23:08 | |
| *** WackoRobie has quit IRC | 23:10 | |
| *** dims has joined #openstack-neutron | 23:26 | |
| *** banix has quit IRC | 23:29 | |
| *** thedodd has quit IRC | 23:37 | |
| *** Jabadia has joined #openstack-neutron | 23:46 | |
| *** Jabadia has quit IRC | 23:51 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
