*** k_mouza has joined #openstack-kolla | 00:00 | |
*** k_mouza has quit IRC | 00:05 | |
*** LinPeiWen46 has joined #openstack-kolla | 00:36 | |
*** ricolin has joined #openstack-kolla | 00:37 | |
openstackgerrit | Lin PeiWen proposed openstack/kolla-ansible stable/victoria: Add support for GlusterFS NFS Manila backend https://review.opendev.org/763292 | 00:43 |
---|---|---|
*** wuchunyang has joined #openstack-kolla | 00:56 | |
*** zzzeek has quit IRC | 02:17 | |
*** zzzeek has joined #openstack-kolla | 02:19 | |
*** jobewan has joined #openstack-kolla | 02:27 | |
*** ivan_lin has joined #openstack-kolla | 02:30 | |
*** LinPeiWen has quit IRC | 02:32 | |
*** zzzeek has quit IRC | 02:35 | |
*** zzzeek has joined #openstack-kolla | 02:36 | |
*** skramaja has joined #openstack-kolla | 03:10 | |
*** bsanjeewa has joined #openstack-kolla | 03:16 | |
*** bsanjeewa has quit IRC | 03:38 | |
*** imtiazc has quit IRC | 04:01 | |
*** wuchunyang has quit IRC | 04:06 | |
*** kevko has quit IRC | 04:11 | |
*** vishalmanchanda has joined #openstack-kolla | 04:30 | |
*** jaicaa has quit IRC | 04:56 | |
*** jaicaa has joined #openstack-kolla | 04:56 | |
openstackgerrit | likui proposed openstack/kolla-ansible master: Update docs for Ubuntu 20.04 https://review.opendev.org/763167 | 05:16 |
*** evrardjp has quit IRC | 05:33 | |
*** evrardjp has joined #openstack-kolla | 05:33 | |
*** ivan_lin has quit IRC | 05:36 | |
*** ivan_lin has joined #openstack-kolla | 05:36 | |
*** k_mouza has joined #openstack-kolla | 06:01 | |
*** JamesBenson has quit IRC | 06:03 | |
*** JamesBenson has joined #openstack-kolla | 06:04 | |
*** k_mouza has quit IRC | 06:06 | |
*** JamesBenson has quit IRC | 06:09 | |
*** JamesBenson has joined #openstack-kolla | 06:15 | |
*** kevko has joined #openstack-kolla | 06:33 | |
*** jaicaa has quit IRC | 06:37 | |
*** jaicaa has joined #openstack-kolla | 06:39 | |
*** cah_link has joined #openstack-kolla | 07:02 | |
*** wuchunyang has joined #openstack-kolla | 07:08 | |
*** bhagyashris|off is now known as bhagyashris | 07:55 | |
*** nikparasyr has joined #openstack-kolla | 07:59 | |
*** rpittau|afk is now known as rpittau | 08:03 | |
mnasiadka | morning | 08:27 |
*** bengates has joined #openstack-kolla | 08:37 | |
*** imcsk8 has quit IRC | 08:37 | |
*** imcsk8 has joined #openstack-kolla | 08:39 | |
*** gfidente|afk is now known as gfidente | 08:46 | |
*** cah_link1 has joined #openstack-kolla | 08:51 | |
*** cah_link has quit IRC | 08:53 | |
*** cah_link1 is now known as cah_link | 08:53 | |
kevko | hi guys, can someone advise me with execution of role in a loop to bunch of servers in several groups ? | 08:54 |
kevko | i will explain to detail if someone :) .. | 08:54 |
kevko | i really think that ansible can't do this | 08:55 |
*** mgoddard has joined #openstack-kolla | 08:58 | |
*** bsanjeewa has joined #openstack-kolla | 09:11 | |
bsanjeewa | mgoddard: about http://paste.openstack.org/show/800169/ yesterday, now both nodes boot into grub> prompt | 09:12 |
mnasiadka | bsanjeewa: yeah, that's something wrong with DIB (or it's elements), if you type exit in the grub> prompt, you'll get to a menu | 09:16 |
mnasiadka | bsanjeewa: running grub2-install on a booted OS seems to fix things | 09:16 |
*** bsanjeewa_ has joined #openstack-kolla | 09:18 | |
bsanjeewa_ | mnasiadka: When type exit on grub>, StartImage failed: Load Error displayed and system does not respond anymore. Will try to locate the kernal and ramdisk images and try the boot command | 09:20 |
mnasiadka | bsanjeewa_: oh, that's probably some different problem then | 09:20 |
*** bsanjeewa has quit IRC | 09:20 | |
*** skramaja has quit IRC | 09:29 | |
osmanlicilegi | morning | 09:29 |
*** bsanjeewa_ has quit IRC | 09:36 | |
*** ricolin has quit IRC | 09:38 | |
openstackgerrit | wu.shiming proposed openstack/kayobe master: Update TOX_CONSTRAINTS_FILE https://review.opendev.org/763329 | 09:39 |
openstackgerrit | Mark Goddard proposed openstack/kayobe master: Update TOX_CONSTRAINTS_FILE https://review.opendev.org/763329 | 09:48 |
mgoddard | morning | 09:48 |
*** bsanjeewa has joined #openstack-kolla | 09:53 | |
bsanjeewa | mnasiadka: the /boot/efi folder is empty. will not boot with kernel in the /boot folder. What should I do? Any additional settings in ipa.yml and bifrost.yml? | 09:55 |
*** k_mouza has joined #openstack-kolla | 09:57 | |
openstackgerrit | wu.chunyang proposed openstack/kolla-ansible master: [CI] Add trove smoke test CI https://review.opendev.org/763197 | 10:12 |
openstackgerrit | Mark Goddard proposed openstack/kolla master: docs: add code reviews to contributor guide https://review.opendev.org/762817 | 10:12 |
*** bsanjeewa_ has joined #openstack-kolla | 10:15 | |
*** wuchunyang has quit IRC | 10:17 | |
*** bsanjeewa has quit IRC | 10:18 | |
*** luksky has joined #openstack-kolla | 10:18 | |
openstackgerrit | Mark Goddard proposed openstack/kayobe-config master: Update TOX_CONSTRAINTS_FILE https://review.opendev.org/763338 | 10:23 |
openstackgerrit | Mark Goddard proposed openstack/kayobe-config-dev master: Update TOX_CONSTRAINTS_FILE https://review.opendev.org/763339 | 10:24 |
*** dardelean has joined #openstack-kolla | 10:24 | |
openstackgerrit | Merged openstack/kayobe master: Update master for stable/victoria https://review.opendev.org/763221 | 10:28 |
*** bsanjeewa_ has quit IRC | 10:28 | |
*** e0ne has joined #openstack-kolla | 10:29 | |
openstackgerrit | Michal Nasiadka proposed openstack/kolla master: CI: Improve publish playbook https://review.opendev.org/763146 | 10:34 |
mnasiadka | hrw: commit message updated - happier? :) | 10:35 |
openstackgerrit | Merged openstack/kolla-ansible stable/victoria: kibana: Remove 6.x migration from upgrade https://review.opendev.org/763214 | 10:36 |
hrw | +2 | 10:36 |
openstackgerrit | Merged openstack/kolla-ansible master: Fix kolla-ansible to work with pyenv-virtualenv https://review.opendev.org/762365 | 10:42 |
openstackgerrit | Mark Goddard proposed openstack/kolla master: Remove footer block from intermediate images https://review.opendev.org/763027 | 11:03 |
openstackgerrit | Mark Goddard proposed openstack/kolla master: WIP: CI: revert to public package mirrors after build https://review.opendev.org/761928 | 11:03 |
openstackgerrit | Mark Goddard proposed openstack/kolla master: WIP: CI: revert to public package mirrors after build https://review.opendev.org/761928 | 11:04 |
openstackgerrit | Mark Goddard proposed openstack/kolla master: WIP: CI: revert to public package mirrors after build https://review.opendev.org/761928 | 11:07 |
kevko | anyone for little discussion about kolla-ansible with me please ? :) | 11:11 |
openstackgerrit | Mark Goddard proposed openstack/kolla master: DNM: testing mirror revert https://review.opendev.org/763351 | 11:11 |
*** cah_link1 has joined #openstack-kolla | 11:13 | |
rockey | kevko: post your question and perhaps someone will have/take the time to answer | 11:15 |
*** cah_link has quit IRC | 11:15 | |
*** cah_link1 is now known as cah_link | 11:15 | |
kevko | ok, i would like to implement mysql proxying and database sharding ... we have it on production and working very good, as shard-router and proxy we are using maxscale | 11:17 |
kevko | my problem is that i want to write maxscale role , but want to also "recycle" mariadb role | 11:17 |
kevko | but if I add servers , or groups to mariadb group in hosts file , site.yml is applying changes to all servers , actually i really need to apply some role to N groups of mariadb clusters | 11:18 |
kevko | this can be done by running playbook with limit ... or playbook to specific group .. so i can't include it to site.yml | 11:19 |
kevko | second option is rewrite mariadb role to do both things ..but it is quite hard .. :/ don't know how to deal with it | 11:19 |
kevko | hope, that it is understandable ... :/ | 11:20 |
mnasiadka | so, you need to run the maxscale role after the mariadb role, and then it needs to run the mariadb role again? | 11:24 |
kevko | maybe I explain problem which we were trying to solve .. | 11:34 |
kevko | kolla is installing mariadb on some hosts right ? (mariadb group, and to its children , by default controller) | 11:35 |
kevko | every database is after this installed to this cluster, but haproxy is sending always to one server, other servers are backups , right ? | 11:36 |
openstackgerrit | Bartosz Bezak proposed openstack/kayobe master: Support for firewalld ZONE option in network interfaces configuration https://review.opendev.org/763180 | 11:37 |
kevko | ^^ this means that most big sense is to have always 2 servers of mariadb, if you have 3,4,N servers, all trafic will be router always to one | 11:37 |
kevko | ^^ this basically means that in some point of cloud usage this one server is not doing his work good ..IO, cpus, network ..etc | 11:38 |
kevko | (especially neutron DB has big ammount of connections) | 11:39 |
kevko | ^^ So , i want to replace haproxy and place maxscale, which have shardRouter on entrypoint ..and sending all trafic per shard (openstack database_name) to group of mariadb | 11:40 |
*** mushrushu has quit IRC | 11:40 | |
kevko | so, firstly i would like to define SOMEHOW ... install mariadb1_cluster here, mariadb_cluster2 here, mariadb_cluster3 here ..and so on ... and after all clusters(shards) is completly built ..i just deploy maxscale container with all backends configured on controller | 11:41 |
kevko | point is , that i really don't know how to install set of mariadb clusters ...because mariadb role is applying her tasks to all servers :/ | 11:42 |
kevko | is it more understandable now :/ ? | 11:42 |
*** mushrushu has joined #openstack-kolla | 11:43 | |
kevko | in next step i just set shard for openstack service ..and bootstrap of mysql_database will go to specific shard ..while all databases will be accessible via maxscale | 11:44 |
*** bsanjeewa_ has joined #openstack-kolla | 11:46 | |
*** bsanjeewa_ has quit IRC | 11:53 | |
*** sluna has quit IRC | 12:10 | |
*** sluna has joined #openstack-kolla | 12:10 | |
mnasiadka | kevko: we understand the motivation to use maxscale, you don't need to explain :) | 12:11 |
openstackgerrit | Merged openstack/kayobe master: Add Python3 wallaby unit tests https://review.opendev.org/763222 | 12:11 |
openstackgerrit | Merged openstack/kayobe-config stable/victoria: Update .gitreview for stable/victoria https://review.opendev.org/763223 | 12:11 |
openstackgerrit | Merged openstack/kayobe-config stable/victoria: Update TOX_CONSTRAINTS_FILE for stable/victoria https://review.opendev.org/763224 | 12:11 |
mnasiadka | kevko: but I don't understand why do you want to change mariadb role, it's deploying a Galera cluster on a given group of nodes (group in Ansible inventory), right? | 12:12 |
openstackgerrit | Merged openstack/kayobe-config-dev stable/victoria: Update .gitreview for stable/victoria https://review.opendev.org/763225 | 12:15 |
openstackgerrit | Merged openstack/kayobe-config-dev stable/victoria: Update TOX_CONSTRAINTS_FILE for stable/victoria https://review.opendev.org/763226 | 12:15 |
*** ricolin has joined #openstack-kolla | 12:18 | |
mnasiadka | kevko: looking at https://mariadb.com/kb/en/mariadb-maxscale-20-maxscale-connection-routing-with-galera-cluster/ - it doesn't seem like you need to do anything with the mariadb role | 12:20 |
*** kevinz has joined #openstack-kolla | 12:36 | |
openstackgerrit | Mark Goddard proposed openstack/kolla master: DNM: testing mirror revert https://review.opendev.org/763351 | 12:36 |
*** zzzeek has quit IRC | 12:45 | |
*** zzzeek has joined #openstack-kolla | 12:47 | |
openstackgerrit | Victor Chembaev proposed openstack/kolla-ansible master: RabbitMQ handler refactored to restart services in serial https://review.opendev.org/763137 | 12:49 |
kevko | mnasiadka: well, imagine then i want to configure neutron database to run on server1,server2 galera cluster, nova on server3,server4 galera cluster right ? | 12:54 |
kevko | but i want to have VIP:3306 to see both clusters | 12:55 |
kevko | *both databases | 12:55 |
kevko | i can run: | 12:57 |
*** hrw has quit IRC | 12:57 | |
kevko | kolla-ansible -i inventory deploy -t mariadb --limit first_group | 12:57 |
kevko | kolla-ansible -i inventory deploy -t mariadb --limit second_group | 12:57 |
kevko | and then kolla-ansible -i inventory deploy | 12:57 |
kevko | point is ..that i have to do it in separate bunch of playbooks to "prepare" environment .. | 12:58 |
*** hrw has joined #openstack-kolla | 12:59 | |
*** cah_link1 has joined #openstack-kolla | 13:03 | |
*** cah_link has quit IRC | 13:05 | |
*** cah_link1 is now known as cah_link | 13:05 | |
*** jbadiapa has joined #openstack-kolla | 13:09 | |
*** rpittau is now known as rpittau|brb | 13:16 | |
openstackgerrit | Radosław Piliszek proposed openstack/kolla-ansible stable/victoria: Fix kolla-ansible to work with pyenv-virtualenv https://review.opendev.org/763366 | 13:34 |
openstackgerrit | Radosław Piliszek proposed openstack/kolla-ansible stable/ussuri: Fix kolla-ansible to work with pyenv-virtualenv https://review.opendev.org/763367 | 13:34 |
openstackgerrit | Radosław Piliszek proposed openstack/kolla-ansible stable/train: Fix kolla-ansible to work with pyenv-virtualenv https://review.opendev.org/763368 | 13:34 |
openstackgerrit | Radosław Piliszek proposed openstack/kolla-ansible stable/stein: Fix kolla-ansible to work with pyenv-virtualenv https://review.opendev.org/763369 | 13:34 |
yoctozepto | kevko: one thing to notice - you probably still want 3 servers per shard - to avoid a situation where you lose quorum | 13:39 |
yoctozepto | kevko: I would say let's do it in small steps | 13:40 |
yoctozepto | kevko: sharding can be added later | 13:40 |
kevko | yoctozepto: ok, doesn't matter if it will be 2, 3, or 4 servers .. | 13:40 |
kevko | yoctozepto: but do you understand what I mean ? | 13:40 |
yoctozepto | kevko: kind of, you want to divide them into shards | 13:40 |
kevko | yoctozepto: well, yes, maxscale is very simple to write .. | 13:40 |
yoctozepto | kevko: what I'm saying is maxscale without shards would be a great boon already | 13:40 |
yoctozepto | kevko: and then we can think about how to approach sharding | 13:41 |
kevko | yoctozepto: yeah, that's true .. | 13:41 |
yoctozepto | kevko: we would likely want some logic to that | 13:41 |
yoctozepto | kevko: I've never run sharding in sql databases | 13:41 |
yoctozepto | only ever mongodb afair :D | 13:41 |
yoctozepto | so would need to read up on that | 13:41 |
yoctozepto | no idea about others | 13:41 |
kevko | yoctozepto: especially group_by is hell in mariadb :D | 13:42 |
yoctozepto | kevko: :D | 13:42 |
kevko | i mean that group_by in lookup_cluster.yml | 13:43 |
kevko | that is reaaaally depresing | 13:43 |
yoctozepto | kevko: well, it's worky | 13:44 |
yoctozepto | and better than what it was :-) | 13:44 |
yoctozepto | always open to improvements | 13:45 |
kevko | yoctozepto: agree, it's mostly ansible limitations ..i know | 13:45 |
openstackgerrit | Merged openstack/kolla master: CI: Improve publish playbook https://review.opendev.org/763146 | 13:46 |
mnasiadka | kevko: simple baby steps, we don't support multiple galera clusters today, so it's totally different feature :) | 13:46 |
kevko | ok, understand .. baby steps :) | 13:47 |
mgoddard | kevko: this was started for nova cells: https://review.opendev.org/619756 | 13:49 |
patchbot | patch 619756 - kolla-ansible - Minimal support for deploying multiple instances o... - 19 patch sets | 13:49 |
kevko | mgoddard: haha, it looks like author was doing steps similar as me :) | 13:51 |
mgoddard | kevko: but could you achieve what you want with a single galera cluster, but different active nodes for different DBs? | 13:52 |
mgoddard | I haven't used maxscale | 13:52 |
kevko | mgoddard: well, simply said ..on one DB galera cluster you have different databases .. right ? | 13:54 |
mgoddard | kevko: yes | 13:54 |
kevko | mgoddard: for example glance is using database very few ... on other side ..there is neutron DB which has on big Openstack clusters very lot connections ...and performance is going down | 13:56 |
mgoddard | kevko: yes, I understand | 13:56 |
kevko | on customer side we have 100+ compute nodes .. | 13:56 |
kevko | and we measured that neutron has problem with DB sometimes .. | 13:56 |
kevko | so, we build another cluster on different set of barememtals | 13:56 |
kevko | and placed neutron database to it .. | 13:57 |
mnasiadka | ML2/OVS? | 13:57 |
mgoddard | kevko: as you said earlier, the issue is that haproxy uses one active backend for all DBs. Could different DBs use different active backends? | 13:57 |
openstackgerrit | Mark Goddard proposed openstack/kayobe master: Revert "Use OpenStack Victoria release" https://review.opendev.org/763374 | 13:58 |
kevko | it's not issue , same functionality can fullfill maxscale .. but mascale can also do sharding ... | 13:58 |
yoctozepto | kevko: that's why we would want to add maxscale as the first step | 13:58 |
kevko | to be honest ..openstack don't like active-active cluster as it is in kolla ...because second replica isn't always fresh | 13:59 |
kevko | so, active-backup has to be as it is now .. | 13:59 |
kevko | yes, you are right ...firstly i should provide simple maxcale switch ..and in second step maybe sharding :) | 14:00 |
openstackgerrit | Merged openstack/kayobe stable/stein: Fix filtering of network names set to an empty string https://review.opendev.org/761332 | 14:01 |
openstackgerrit | Mark Goddard proposed openstack/kayobe master: Update TOX_CONSTRAINTS_FILE https://review.opendev.org/763329 | 14:02 |
yoctozepto | kevko: currently it is active-active behind the scenes (haproxy) | 14:02 |
yoctozepto | because it's multimaster | 14:02 |
kevko | yes, i really know how it is in kolla .. | 14:02 |
kevko | but did you try to turn on active-active in haproxy ? :) | 14:03 |
kevko | i mean, loadbalance transactions to mysql backends ? | 14:03 |
kevko | there is always small delay between replicas .. | 14:04 |
openstackgerrit | Mark Goddard proposed openstack/kayobe master: CI: set previous_release to Victoria https://review.opendev.org/763375 | 14:04 |
*** Arador has joined #openstack-kolla | 14:11 | |
*** TrevorV has joined #openstack-kolla | 14:16 | |
openstackgerrit | Bartosz Bezak proposed openstack/kayobe master: Support for firewalld ZONE option in network interfaces configuration https://review.opendev.org/763180 | 14:20 |
mgoddard | kevko, yoctozepto: I think you can specify different URLs for read vs. write. read being load balanced, write being active/backup | 14:20 |
kevko | mgoddard: and that's problem ...because sometimes it is not synced ..and you want to read .. openstack service fail | 14:22 |
kevko | like, there is only small latency between ..but OS services are very sensitive on that | 14:22 |
kevko | in that setup you can be sure that data is only on first node in every time .. | 14:23 |
kevko | that's the reason why i'm more familiar with ndb-cluster | 14:23 |
openstackgerrit | Mark Goddard proposed openstack/kolla master: DNM: testing mirror revert https://review.opendev.org/763351 | 14:35 |
*** happyhemant has joined #openstack-kolla | 14:44 | |
*** mgoddard has quit IRC | 14:48 | |
*** rpittau|brb is now known as rpittau | 14:51 | |
openstackgerrit | Merged openstack/kayobe stable/victoria: Update .gitreview for stable/victoria https://review.opendev.org/763219 | 15:02 |
openstackgerrit | Merged openstack/kayobe stable/victoria: Update TOX_CONSTRAINTS_FILE for stable/victoria https://review.opendev.org/763220 | 15:02 |
*** mgoddard has joined #openstack-kolla | 15:07 | |
*** k_mouza has quit IRC | 15:15 | |
*** bsanjeewa has joined #openstack-kolla | 15:20 | |
*** k_mouza has joined #openstack-kolla | 15:28 | |
openstackgerrit | Mark Goddard proposed openstack/kolla master: WIP: CI: revert to public package mirrors after build https://review.opendev.org/761928 | 15:37 |
openstackgerrit | Mark Goddard proposed openstack/kolla master: DNM: testing mirror revert https://review.opendev.org/763351 | 15:37 |
*** vishalmanchanda has quit IRC | 16:00 | |
*** diurnalist has joined #openstack-kolla | 16:03 | |
*** diurnalist has quit IRC | 16:04 | |
*** diurnalist has joined #openstack-kolla | 16:04 | |
*** evrardjp has quit IRC | 16:17 | |
Arador | I am getting stuck on "wait for MariaDB to be available via HAProxy" and I am unable to find a working solution online, does anyone have a suggestion? | 16:24 |
Arador | multinode on VM's, installing ussuri on centos 8.2 | 16:26 |
*** nikparasyr has left #openstack-kolla | 16:32 | |
*** rpittau is now known as rpittau|afk | 16:52 | |
*** sluna has quit IRC | 16:53 | |
*** bengates has quit IRC | 16:57 | |
*** bengates has joined #openstack-kolla | 16:58 | |
*** bengates has quit IRC | 17:02 | |
*** happyhemant has quit IRC | 17:03 | |
openstackgerrit | Victor Chembaev proposed openstack/kolla-ansible master: RabbitMQ handler refactored to restart services in serial https://review.opendev.org/763137 | 17:12 |
*** bsanjeewa has quit IRC | 17:30 | |
*** JamesBenson has quit IRC | 17:33 | |
*** diurnalist has quit IRC | 17:46 | |
*** gfidente is now known as gfidente|afk | 17:49 | |
*** cah_link has quit IRC | 17:49 | |
*** mgoddard has quit IRC | 18:05 | |
*** k_mouza has quit IRC | 18:34 | |
*** dardelean has quit IRC | 18:50 | |
*** dciabrin has quit IRC | 19:05 | |
*** mgoddard has joined #openstack-kolla | 19:27 | |
*** mgoddard has quit IRC | 19:39 | |
*** dciabrin has joined #openstack-kolla | 20:02 | |
*** ysirndjuro has quit IRC | 20:10 | |
*** bsanjeewa has joined #openstack-kolla | 20:16 | |
*** bsanjeewa has quit IRC | 20:21 | |
*** luksky has quit IRC | 20:26 | |
*** luksky has joined #openstack-kolla | 20:26 | |
*** ysirndjuro has joined #openstack-kolla | 20:32 | |
*** k_mouza has joined #openstack-kolla | 20:35 | |
*** k_mouza has quit IRC | 20:39 | |
*** jbadiapa has quit IRC | 20:52 | |
*** TrevorV has quit IRC | 21:09 | |
*** dciabrin has quit IRC | 21:11 | |
*** dciabrin has joined #openstack-kolla | 21:11 | |
*** stingrayza has quit IRC | 21:11 | |
*** stingrayza has joined #openstack-kolla | 21:13 | |
*** e0ne has quit IRC | 21:24 | |
*** luksky has quit IRC | 21:41 | |
*** luksky has joined #openstack-kolla | 21:54 | |
*** kevko has quit IRC | 22:37 | |
*** Arador has quit IRC | 23:12 | |
*** strobert has quit IRC | 23:15 | |
*** strobert has joined #openstack-kolla | 23:16 | |
*** kevko has joined #openstack-kolla | 23:27 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!