Perl 6 - the future is here, just unevenly distributed

IRC log for #puppet-openstack, 2014-03-17

| Channels | #puppet-openstack index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:09 thuc joined #puppet-openstack
00:19 ryanycoleman joined #puppet-openstack
00:49 thuc joined #puppet-openstack
00:53 thuc joined #puppet-openstack
01:12 kitp joined #puppet-openstack
01:26 thuc_ joined #puppet-openstack
01:32 thuc joined #puppet-openstack
01:47 thuc joined #puppet-openstack
01:52 ryanycoleman joined #puppet-openstack
01:54 thuc joined #puppet-openstack
02:09 openstackgerrit Jacques Uber proposed a change to stackforge/puppet-openstack: fix broken anchor tag  https://review.openstack.org/80877
02:09 thuc joined #puppet-openstack
02:12 kitp joined #puppet-openstack
02:27 thuc joined #puppet-openstack
02:28 thuc_ joined #puppet-openstack
02:30 thuc_ joined #puppet-openstack
03:13 kitp joined #puppet-openstack
03:51 thuc joined #puppet-openstack
04:00 kitp joined #puppet-openstack
04:01 kitp joined #puppet-openstack
04:29 EmilienM good morning here
04:36 uberj joined #puppet-openstack
05:02 kitp joined #puppet-openstack
05:40 michchap_ EmilienM: morning
05:40 michchap_ EmilienM: are you using the nova::cells code?
05:52 ryanycoleman joined #puppet-openstack
06:03 kitp joined #puppet-openstack
06:27 xarses joined #puppet-openstack
06:47 saju_m joined #puppet-openstack
06:49 namredips joined #puppet-openstack
07:04 kitp joined #puppet-openstack
07:08 dgollub joined #puppet-openstack
07:21 rongze joined #puppet-openstack
07:29 EmilienM michchap_: not yet, why?
07:30 openstackgerrit Benedikt Trefzer proposed a change to stackforge/puppet-neutron: add parameter for endpoint protocols  https://review.openstack.org/80917
07:31 michchap_ EmilienM: you authored it 6 months ago. I tried it out today and it doesn't work for me, I wanted to see if you were using it or if I can rewrite it without worrying about backwards compatibility
07:31 EmilienM michchap_: there is something I needed to fix
07:31 EmilienM it's the order
07:31 EmilienM you have to 1) configure the nova.conf 2) run the nova-manage 3) start the service
07:32 EmilienM and then it should work (tested)
07:33 openstackgerrit Benedikt Trefzer proposed a change to stackforge/puppet-cinder: add parameter for endpoint protocols  https://review.openstack.org/80918
07:33 michchap_ there is a call_parent_name parameter being passed from nova::manage::cells to nova_cells that doesn't exist, and the regex for matching cell names in the provider wasn't working for me either
07:33 mkoderer joined #puppet-openstack
07:34 EmilienM michchap_: ouch. Feel free to fix it
07:34 michchap_ the other thing is I added a hash parameter for specifying cells without using storeconfigs
07:34 michchap_ EmilienM: will do. Just wanted to see if you'd tested it since Grizzly
07:35 EmilienM michchap_: just on PoC, not really in production
07:35 EmilienM michchap_: cells stuffs are not really ready for production when using neutron
07:35 michchap_ EmilienM: np, we're about to push it out to a customer site so we're putting it through some testing
07:35 EmilienM michchap_: cool
07:35 michchap_ EmilienM: what doesn't work with Neutron + cells?
07:36 EmilienM michchap_: it works, but it's useless
07:36 EmilienM michchap_: cells aims to provide high scalability for compute, but without network celling you loose the compute scalability
07:37 michchap_ EmilienM: so you can't have one neutron per cell, it has to be shared?
07:37 EmilienM so even though your VM is schedule on Cell #2, you will probably have your network scheduled far away
07:40 michchap_ so you can have many compute cells but they must all talk to one neutron? as opposed to nova network where each cell has their own nova-network
07:41 EmilienM michchap_: if you have multi host enabled on nova network, that's fine
07:41 EmilienM because l3 is handled by compute nodes
07:42 EmilienM michchap_: the fact is neutron is not cells-aware at all
07:42 michchap_ EmilienM: so this is just an l3 problem (as usual)
07:42 EmilienM michchap_: yup but not only
07:43 EmilienM also a L2, because you can have a lot of GRE tunnels between compute nodes across cells while you certainly would prefer to have them within one cell
07:43 michchap_ so each cell does not have its own Neutron DB then?
07:43 EmilienM michchap_: nop
07:44 EmilienM michchap_: keep in mind, only nova is cells-aware today
07:44 michchap_ EmilienM: so the cells have to be on the same l2 network for GRE to work
07:44 michchap_ EmilienM: yeah, I was just wondering how it works
07:44 michchap_ EmilienM: I'm probably going to use provider networks for this specific site, until there is a better solution
07:45 EmilienM michchap_: not really need, the GRE are built over IP so the tunnel interfaces must have l3 connectivity between them
07:46 michchap_ EmilienM: they have to be able to route to each other
07:46 EmilienM michchap_: yep, a ping test is enough
07:46 michchap_ EmilienM: thanks for the info.
07:46 EmilienM and no firewall rule between them which would block GRE protocol
07:47 EmilienM (if you use GRE)
07:47 EmilienM michchap_: i would suggest using VXLAN by the way
07:48 michchap_ EmilienM: We're using Mellanox plugin, it has its own L2 thing
07:48 EmilienM michchap_: wow, you should have awesome perfs :)
07:49 michchap_ EmilienM: yep they are already getting very good results - ~45Gb/s VM to VM
07:54 EmilienM michchap_: omg
07:54 EmilienM michchap_: using GRE?
07:55 michchap_ Nah using the mellanox L2 - it's doing SRIOV or something very similar
07:55 michchap_ I haven't got access to the site yet, but I will take a look when I get a login :)
07:56 EmilienM michchap_: ah ok
08:04 kitp joined #puppet-openstack
08:26 mkoderer_ joined #puppet-openstack
08:27 dgollub joined #puppet-openstack
08:29 bauzas joined #puppet-openstack
08:37 fandi joined #puppet-openstack
08:42 EmilienM fvollero: hey I think i found an issue in packstack about heat
08:52 ryanycoleman joined #puppet-openstack
09:05 kitp joined #puppet-openstack
09:13 EmilienM michchap_: still around?
09:48 fandi joined #puppet-openstack
10:02 derekh joined #puppet-openstack
10:04 michchap_ EmilienM: yep
10:06 kitp joined #puppet-openstack
10:11 BroadcastStorm joined #puppet-openstack
10:16 mmagr joined #puppet-openstack
10:26 dgollub joined #puppet-openstack
10:32 saju_m joined #puppet-openstack
10:52 fvollero EmilienM: Shoot man
10:56 openstackgerrit A change was merged to stackforge/puppet-vswitch: Fix openvswitch-switch status command on Ubuntu platform  https://review.openstack.org/78920
10:57 EmilienM fvollero: heat-engine should be run ONCE
10:57 EmilienM if you see what I mean
11:00 fvollero EmilienM: Oh, it's executed more than once?
11:00 fvollero EmilienM: that's totally weird, gonna check it
11:00 EmilienM fvollero: if you run your heat manifest on multi nodes, yeah :)
11:03 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-keystone: ensure log_file is absent when using syslog  https://review.openstack.org/80633
11:04 EmilienM sbadia: if you got time, you could review that one ^
11:04 EmilienM fvollero: i don't know if you can run packstack with multiple heat servers though
11:04 EmilienM if it's the case, you're done :)
11:06 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-glance: Update glance_image provider with API v2  https://review.openstack.org/67768
11:07 kitp joined #puppet-openstack
11:11 openstackgerrit joined #puppet-openstack
11:24 fvollero EmilienM: should not be the "default" :)
11:34 beddari mattymo: got any current info on running ceph on compute nodes? :) I'm seeing a need for this … for some workloads I just plain need local storage and thus it seems the path of least resistance
11:35 mattymo beddari, we have it working in Fuel
11:35 mattymo but I didn't write any code related to it, unfortunately
11:35 mattymo xarses does
11:35 beddari mattymo: mmm .. I'm thinking I'd try it "carefully" first, but do you use it as a general default?
11:36 saju_m joined #puppet-openstack
11:36 mattymo no we default to cinder with lvm backend and cinder to file backend
11:36 mattymo oops glance to file backend
11:36 britthouser joined #puppet-openstack
11:36 beddari yea
11:36 mattymo ceph requires a lot of memory and cpu
11:36 mattymo enough so that when 1 node drops offline, it can hammer the CPU for 5-10 minutes
11:37 beddari mm, I'm hoping it could work for databases but this is the being careful part, yes
11:38 mattymo so you need to budget 1 cpu core for just ceph
11:45 openstackgerrit A change was merged to stackforge/puppet_openstack_builder: Fix outdated openstack_repo_location for HA  https://review.openstack.org/80362
11:59 dgollub joined #puppet-openstack
12:00 openstackgerrit A change was merged to stackforge/puppet-ceilometer: Changes file_path to self.file_path  https://review.openstack.org/73620
12:01 ryanycoleman joined #puppet-openstack
12:03 dprince joined #puppet-openstack
12:07 kitp joined #puppet-openstack
12:10 dgollub joined #puppet-openstack
12:19 rcrit left #puppet-openstack
12:21 markvoelker joined #puppet-openstack
13:01 ryanycoleman joined #puppet-openstack
13:05 EmilienM michchap_: i had this conversation with hogepodge about the readme in https://review.openstack.org/#/c/80070/
13:05 EmilienM michchap_: and i don't know what to do in that file
13:07 michchap_ EmilienM: what do you mean? The patch is about multi-backend, so I think just leave the readme alone unless you want to add a line about multi-backend
13:08 EmilienM michchap_: it's a backport
13:08 EmilienM just a cherry pick
13:08 kitp joined #puppet-openstack
13:09 michchap_ EmilienM: yep
13:09 michchap_ EmilienM: the original is here? https://github.com/stackforge/puppet-cinder/commit/209a5bd36af085cf306cf084ed1af1bf0fc9f41d
13:10 michchap_ EmilienM: you just need to stop the release notes section from being overwritten, everything else should stay the same as that patch. Or am I not understanding?
13:10 EmilienM michchap_: oh, ok
13:10 EmilienM let's try again then :)
13:12 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-cinder: [bp] Add support for Cinder multi-backend  https://review.openstack.org/80070
13:12 EmilienM michchap_: ^
13:12 thuc joined #puppet-openstack
13:12 EmilienM oops
13:12 EmilienM not good
13:13 michchap_ ah we lost the 3.0.0 section
13:13 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-cinder: [bp] Add support for Cinder multi-backend  https://review.openstack.org/80070
13:13 EmilienM should be good ^
13:13 thuc_ joined #puppet-openstack
13:13 EmilienM michchap_: ^
13:14 michchap_ EmilienM: :)
13:14 EmilienM michchap_: thanks for the help, i'm finally get through that one
13:14 michchap_ EmilienM: oh it's THIS patch :o
13:15 EmilienM michchap_: :P
13:15 michchap_ EmilienM: I regretted merging this about 30 minutes after I approved
13:15 michchap_ EmilienM: because there's duplicate code everywhere now for the backends
13:18 michchap_ EmilienM: should we deprecate the non-multi-backend case? so that everything is behind backends instead of being both volume/ and backend/? What do you think?
13:24 EmilienM michchap_: no, we should not deprecate the non multi backend case I think
13:25 EmilienM michchap_: in your eyes it's dupplicated but actually it's not when you look at how it configures cinder.
13:25 EmilienM michchap_: the params are copy / pasted, agree, not the way to configure
13:26 michchap_ EmilienM: If I configure multi-backend using only one backend, how is that different from using volume in the way cinder is configured?
13:26 EmilienM michchap_: you have to create a volume type
13:26 EmilienM i agree the result is almost the same
13:27 michchap_ EmilienM: so client requests must specify type?
13:27 EmilienM yep
13:27 michchap_ EmilienM: or you can set a default?
13:27 EmilienM not sure, let me check
13:27 michchap_ EmilienM: sounds a little bit like the case with ML2
13:28 EmilienM michchap_: the type is mandatory
13:28 michchap_ EmilienM: ah, so it's not backwards compatible, that's sad
13:29 EmilienM michchap_: yep :)
13:29 EmilienM michchap_: what is bad, is that the feature has been here since grizzly and we added it a few time ago only :(
13:30 michchap_ EmilienM: yeah. We are very far behind for some of the projects.
13:30 EmilienM yep
13:30 EmilienM that's why i'm backporting
13:31 EmilienM because people should use it since the first release they deploy OpenStack
13:31 EmilienM michchap_: my scenario is that people starts with havana, so they will start with good practises (multi backend in cinder for example), and the migration would be easy
13:32 EmilienM michchap_: oh, good news: default_volume_type in cinder.conf
13:32 michchap_ EmilienM: added in icehouse?
13:32 EmilienM http://openstack-in-production.blogspot.fr/2014/03/enable-cinder-multi-backend-with.html
13:33 EmilienM michchap_: no
13:33 michchap_ EmilienM: ah havana
13:33 michchap_ EmilienM: yay
13:33 EmilienM michchap_: :)
13:35 EmilienM michchap_: i'm about adding in puppet-cinder, where would you see it ? in init?
13:35 EmilienM michchap_: or maybe in cinder::backends
13:36 michchap_ backends sounds right
13:37 EmilienM michchap_: the question is: where should this flag put?
13:37 EmilienM I'm checking the code
13:37 dmsimard joined #puppet-openstack
13:38 EmilienM michchap_: on cinder volume nodes
13:38 michchap_ EmilienM: oh? not on the api node?
13:38 EmilienM nop
13:38 michchap_ EmilienM: I have no idea.
13:39 EmilienM michchap_: no i'm sure
13:39 michchap_ dmsimard: Thank you for your blog post about yaml anchors in heira :) It was very helpful. I wish Hiera could use anchors across file boundaries.
13:39 EmilienM volume
13:40 kkirkpatrick joined #puppet-openstack
13:43 dmsimard michchap_: no problem, glad it's useful ! I believe you can use anchors across files but the anchor must be before the reference in the hierarchy
13:43 dmsimard michchap_: I remember trying it, I think being on same hierarchy level is not good enough, the anchor needs to be declared before - hiera doesn't load the whole hierarchy before matching references
13:45 michchap_ dmsimard: I will experiment a bit more...my first tests with cross file were failures, but perhaps I was missing something.
13:45 michchap_ dmsimard: have you seen this? https://github.com/bodepd/scenario_node_terminus#data-mappings
13:46 michchap_ dmsimard: I think hiera could benefit from something built in for 'mappings' like this - interpolation flattens everything to strings so it's unsuitable for a lot of data.
13:47 dmsimard michchap_: yeah.. scenario_node_terminus introduces a lot of cool things.
13:48 dmsimard michchap_: I read somewhere about someone wanting to implement hiera straight within modules - like for default values for example. There's /lib/providers, /lib/puppet/fact for instance, you would have /lib/hiera too.
13:49 michchap_ dmsimard: that's an interesting idea.
13:58 thuc joined #puppet-openstack
13:58 namredips joined #puppet-openstack
14:02 ryanycoleman joined #puppet-openstack
14:07 rwsu joined #puppet-openstack
14:09 kitp joined #puppet-openstack
14:11 thuc joined #puppet-openstack
14:15 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-cinder: multi-backend: configure default type  https://review.openstack.org/80978
14:15 EmilienM michchap_: done^
14:16 michchap_ EmilienM: it will set 'undef' in config by default I think
14:16 EmilienM michchap_: which is bad
14:16 EmilienM michchap_: it could leads to an error
14:16 EmilienM michchap_: if your client does not specify a type
14:18 michchap_ EmilienM: hmm...maybe warn? Is it likely someone wants to deploy with no default storage backend?
14:19 EmilienM michchap_: then he sets undef, no?
14:20 EmilienM michchap_: i can let false by default
14:20 EmilienM if you prefer
14:21 michchap_ EmilienM: I mean don't set the setting at all
14:21 EmilienM hum
14:21 otherwiseguy joined #puppet-openstack
14:21 EmilienM michchap_: so I abandon this patch?
14:22 michchap_ EmilienM: No the patch is good, just if $default_volume_type { ... }
14:22 EmilienM ok, let me show you my next patchset in a few minutes
14:22 michchap_ EmilienM: otherwise I think it converts to a string and literally sets 'false' or 'undef' in the config
14:22 EmilienM ah
14:25 thuc joined #puppet-openstack
14:29 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-cinder: multi-backend: configure default type  https://review.openstack.org/80978
14:29 EmilienM michchap_: ^
14:32 michchap_ EmilienM: you have tabs in that file :D
14:33 EmilienM michchap_: i'm trying a new editor, my bad
14:33 michchap_ EmilienM: oh which one?
14:33 EmilienM michchap_: sublime
14:33 michchap_ EmilienM: good?
14:33 EmilienM michchap_: ~
14:34 EmilienM michchap_: i come back to vi
14:34 EmilienM michchap_: what is yours?
14:34 michchap_ EmilienM: vi
14:34 michchap_ EmilienM: only thing installed on every server
14:34 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-cinder: multi-backend: configure default type  https://review.openstack.org/80978
14:38 michchap_ EmilienM: I think our lint is a bit lax - it should catch those tabs
14:39 michchap_ EmilienM: btw there are still tabs in manfests/backends.pp
14:39 EmilienM damn
14:39 michchap_ EmilienM: I'm off to sleep, I will review tomorrow.
14:39 EmilienM michchap_: where?
14:40 EmilienM michchap_: sure, good night :) thanks for staying around !!
14:40 EmilienM i did not realize it's 1.40 am
14:41 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-cinder: multi-backend: configure default type  https://review.openstack.org/80978
14:43 kkirkpatrick joined #puppet-openstack
14:53 ryanycoleman joined #puppet-openstack
14:55 thuc joined #puppet-openstack
14:58 thuc joined #puppet-openstack
14:59 EmilienM hogepodge: could you approve https://review.openstack.org/#/c/80070/ please?
14:59 thuc_ joined #puppet-openstack
15:02 dmsimard ahah, it's not "can you review", it's "can you approve" :P
15:02 dmsimard suggestive enough :)
15:03 EmilienM dmsimard: yeah, he told me that it should been approved this week-end :)
15:04 dmsimard Hey EmilienM, i'll talk about Ceph and someone from Enovance will talk about swift at a university in Montreal tonight
15:04 EmilienM dmsimard: really? who?
15:04 dmsimard Don't know who, though, doubt it'll be chmouel !?
15:04 dmsimard Would be great to chat with chmouel :)
15:05 dmsimard EmilienM: The only thing I know is that the enovance contact for the event is Marcos Garcia, don't know who will be doing the actual talk
15:05 EmilienM dmsimard: feel free to chat with him on IRC, "chmouel" on freenode
15:06 dmsimard EmilienM: Yeah, I know, but in person - I meant
15:06 EmilienM dmsimard: oh, I see
15:10 kkirkpatrick joined #puppet-openstack
15:13 daneyon joined #puppet-openstack
15:13 daneyon joined #puppet-openstack
15:20 saju_m joined #puppet-openstack
15:24 openstackgerrit Mark T. Voelker proposed a change to stackforge/puppet_openstack_builder: Move to Icehouse  https://review.openstack.org/80739
15:29 otherwiseguy joined #puppet-openstack
15:33 openstackgerrit Mark T. Voelker proposed a change to stackforge/puppet_openstack_builder: Move to Icehouse  https://review.openstack.org/80739
15:48 Aimon joined #puppet-openstack
15:53 marun joined #puppet-openstack
15:55 xarses joined #puppet-openstack
16:17 dachary dmsimard: would you have a minute to review https://review.openstack.org/#/c/74645/ ?
16:23 xarses_ joined #puppet-openstack
16:25 dmsimard dachary: yeah, soon
16:30 rmoe joined #puppet-openstack
16:46 dmsimard dachary: I think usage could be documented in the README, i'm not against letting it go like this and having the README being in another commit, though.
16:46 dmsimard What do you think ?
16:47 kkirkpatrick joined #puppet-openstack
16:55 hogepodge_ joined #puppet-openstack
17:02 ryanycoleman joined #puppet-openstack
17:03 hogepodge joined #puppet-openstack
17:05 hogepodge EmilienM merged
17:06 dmsimard xarses: ^
17:07 gonery Salut mgagne1, do you still plan to review https://review.openstack.org/#/c/67726/ otherwise may I ask someone else to take a look?
17:15 bauzas joined #puppet-openstack
17:23 hogepodge joined #puppet-openstack
17:31 rharrison joined #puppet-openstack
17:32 otherwiseguy joined #puppet-openstack
17:34 angdraug joined #puppet-openstack
17:42 sputnik13 joined #puppet-openstack
17:46 openstackgerrit Donald Talton proposed a change to stackforge/puppet_openstack_builder: Removed unused cinder and glance data mappings.  https://review.openstack.org/80085
18:01 ryanycoleman joined #puppet-openstack
18:14 dgollub joined #puppet-openstack
18:29 PEPPERONIE joined #puppet-openstack
18:35 ryanycoleman joined #puppet-openstack
18:41 dtalton joined #puppet-openstack
18:45 ryanycoleman joined #puppet-openstack
18:49 francois1 joined #puppet-openstack
18:52 mmagr joined #puppet-openstack
19:09 mgagne joined #puppet-openstack
19:11 thuc joined #puppet-openstack
19:17 ryanycoleman joined #puppet-openstack
19:19 thumpba joined #puppet-openstack
19:30 fandi joined #puppet-openstack
19:33 thuc joined #puppet-openstack
19:36 dmsimard dachary, xarses: See my last comment up here ? ^
19:37 xarses dmsimard: which commit?
19:37 dmsimard https://review.openstack.org/#/c/74645/
19:40 thuc_ joined #puppet-openstack
19:43 PEPPERONIE joined #puppet-openstack
19:46 kkirkpatrick joined #puppet-openstack
19:47 mkoderer joined #puppet-openstack
19:56 gonery mgagne: thanks
19:58 dmsimard xarses, dachary: I approved, there are no instructions for mon configuration either - we'll need to update the readme for both mon and osd usage.
19:58 xarses dmsimard: maybe we should log a bug to track
19:59 dmsimard xarses: Sure.
20:00 dmsimard xarses: https://bugs.launchpad.net/puppet-ceph/+bug/1293776
20:03 xarses dmsimard: thanks
20:08 openstackgerrit A change was merged to stackforge/puppet-horizon: improve WSGI configuration  https://review.openstack.org/67726
20:09 openstackgerrit A change was merged to stackforge/puppet-horizon: bump minimal stdlib version to 4.0.0  https://review.openstack.org/80008
20:11 openstackgerrit A change was merged to stackforge/puppet-ceph: define ceph::osd  https://review.openstack.org/74645
20:18 openstackgerrit Donald Talton proposed a change to stackforge/puppet-nova: Added the ability to manage the nova uid/gid.  https://review.openstack.org/76298
20:41 ryanycoleman joined #puppet-openstack
20:57 thuc joined #puppet-openstack
21:04 PEPPERONIE joined #puppet-openstack
21:06 hogepodge ping EmilienM
21:20 dmsimard joined #puppet-openstack
21:27 thuc joined #puppet-openstack
21:31 rwsu joined #puppet-openstack
21:43 PEPPERONIE left #puppet-openstack
21:52 mattymo joined #puppet-openstack
22:10 thuc joined #puppet-openstack
22:11 thuc_ joined #puppet-openstack
22:24 thuc joined #puppet-openstack
22:35 ryanycoleman joined #puppet-openstack
22:40 openstackgerrit joined #puppet-openstack
22:43 hogepodge Any core devs in here right now?
22:50 michchap joined #puppet-openstack
23:03 hogepodge EmilienM you're going to need to rebase https://review.openstack.org/#/c/80071
23:12 openstackgerrit A change was merged to stackforge/puppet-openstack: fix broken anchor tag  https://review.openstack.org/80877
23:17 PEPPERONIE joined #puppet-openstack
23:30 thuc joined #puppet-openstack
23:31 thuc_ joined #puppet-openstack
23:45 thuc joined #puppet-openstack
23:47 thuc_ joined #puppet-openstack

| Channels | #puppet-openstack index | Today | | Search | Google Search | Plain-Text | summary