| *** petertr7_away is now known as petertr7 | 00:02 | |
| mattoliverau | clayg: :( | 00:04 |
|---|---|---|
| mattoliverau | ho: morning | 00:04 |
| *** remix_tj has quit IRC | 00:11 | |
| *** petertr7 is now known as petertr7_away | 00:12 | |
| ho | mattoliverau: morning! | 00:18 |
| openstackgerrit | Darrell Bishop proposed openstack/swift: Allow 1+ object-servers-per-disk deployment https://review.openstack.org/184189 | 00:24 |
| clayg | gooody goody goody | 00:24 |
| swifterdarrell | clayg: new patchset up w/replies inline for Patch Set 13 | 00:25 |
| swifterdarrell | clayg: that should cover everything except the shit in swift-object-server (the hard part) | 00:25 |
| clayg | "Any clusters using threads_per_disk should switch to using servers per port" <- that's calling it like it is boys | 00:30 |
| *** thurloat is now known as thurloat_isgone | 00:32 | |
| *** minwoob has quit IRC | 00:35 | |
| *** haomaiwa_ has quit IRC | 00:43 | |
| *** jrichli has joined #openstack-swift | 00:47 | |
| *** dmorita has joined #openstack-swift | 00:55 | |
| *** amoturi has joined #openstack-swift | 00:56 | |
| openstackgerrit | Tim Burke proposed openstack/python-swiftclient: Add bulkdelete command https://review.openstack.org/190887 | 01:13 |
| *** amoturi has quit IRC | 01:18 | |
| *** annegentle has joined #openstack-swift | 01:26 | |
| *** annegentle has quit IRC | 01:31 | |
| *** BAKfr has quit IRC | 02:31 | |
| *** BAKfr has joined #openstack-swift | 02:37 | |
| mattoliverau | After a few hours of debugging.. and all I needed to add was 'and item_list' to an if statement.. sigh, debugging a cluster can be fun :P #larger_scale_shard_testing | 02:54 |
| mattoliverau | anyway lunch time | 02:54 |
| openstackgerrit | Merged openstack/swift: go: make full disk log message debug-level https://review.openstack.org/190789 | 03:35 |
| *** fifieldt has joined #openstack-swift | 03:42 | |
| *** bill_az has quit IRC | 03:42 | |
| *** evanjfraser_ has joined #openstack-swift | 03:54 | |
| *** evanjfraser has quit IRC | 03:55 | |
| *** jrichli has quit IRC | 04:14 | |
| *** ozialien has joined #openstack-swift | 04:22 | |
| *** tdasilva has quit IRC | 04:32 | |
| *** SkyRocknRoll has joined #openstack-swift | 04:38 | |
| *** ppai has joined #openstack-swift | 04:45 | |
| *** SkyRocknRoll has quit IRC | 04:54 | |
| *** SkyRocknRoll has joined #openstack-swift | 04:55 | |
| *** SkyRocknRoll has joined #openstack-swift | 04:55 | |
| *** tdasilva has joined #openstack-swift | 04:55 | |
| *** SkyRocknRoll has quit IRC | 04:56 | |
| *** SkyRocknRoll has joined #openstack-swift | 04:56 | |
| *** chlong is now known as chlong_mtg | 04:59 | |
| *** zaitcev has quit IRC | 05:09 | |
| ho | acoles: sorry for bother you but please take a look at comments on patch 190370. | 05:10 |
| patchbot | ho: https://review.openstack.org/#/c/190370/ | 05:10 |
| ho | torgomatic: thanks for the comment on above. I think swift can't avoid to introduce oslo config when swift supports policy.json but for swift I read the code it seems just initialize global CONF is enogh. | 05:21 |
| ho | torgomatic: I checked the code when I wrote this: https://review.openstack.org/#/c/149930/14/swift/common/middleware/keystoneauth.py | 05:21 |
| *** proteusguy has quit IRC | 05:24 | |
| *** SkyRocknRoll has quit IRC | 05:29 | |
| ho | torgomatic: swift doesn't use oslo config now. You checked the CONF is global variable but swift doesn't use it and my idea is oslo config will be enabled when keystoneauth (work with openstack components) works. I think the impact to swift is quite small (auth_token and keystoneauth). | 05:55 |
| *** SkyRocknRoll has joined #openstack-swift | 05:58 | |
| *** SkyRocknRoll has joined #openstack-swift | 05:58 | |
| *** chlong_mtg is now known as chlong | 06:06 | |
| *** haomaiwang has joined #openstack-swift | 06:08 | |
| *** ozialien has quit IRC | 06:45 | |
| jith_ | hi all, i m trying to install swift in a VM.. i couldnt understand the command sudo truncate -s 20GB /srv/swift-disk... will it allocate 20GB to swift-disk.. why truncate is used.. i m a beginner in this | 06:47 |
| *** proteusguy has joined #openstack-swift | 07:03 | |
| mattoliverau | jith_: truncate will create a sparse file, so when mounted loop back it can use up to 20GB but only be the size of what is currently used | 07:06 |
| *** mmcardle has joined #openstack-swift | 07:06 | |
| mattoliverau | So don't use all the space now :) | 07:06 |
| jith_ | oh ok thanks mattoliverau :)... how to remove it... i unknowing created 20GB in another VM.. i have to remove it now because the total size allocated for the VM itself is 30GB.. is it enough to do rm -rf /srv/swift-disk.. will it be freed? | 07:10 |
| *** mmcardle1 has joined #openstack-swift | 07:12 | |
| mattoliverau | jith_: yup it's just a file. Unless you've mounted it already, then you have to umount it first | 07:13 |
| *** mmcardle has quit IRC | 07:14 | |
| jith_ | mattoliverau: thanks really... no i havenot mounted.. but whether other folders can use this 20GB if no other space is available?? | 07:16 |
| mattoliverau | jith_: what happens if you using truncate to make a space file, you then put an XFS filesystem on it. You then mount that to a directory and any file or directory placed in that directory is written to the sparse file. It's a way of having an XF partition without having to pass another partition to your VM. | 07:19 |
| mattoliverau | *XFS partition | 07:19 |
| mattoliverau | jith_: so by loop mounting the XFS to the root of the swift folder structure /mnt/<loop_mounted_file>/sdb(1-x) means everything swift writes to the drives (in SAIO) is actally written to that sparse file. So yes they all share it. | 07:23 |
| openstackgerrit | Darrell Bishop proposed openstack/swift: Allow 1+ object-servers-per-disk deployment https://review.openstack.org/184189 | 07:23 |
| ahale | truncate is just nice and fast compared to dd, you dont have to make it 20G, it can be smaller. but you cant use the space for rest of system while the 20G is being filled up | 07:24 |
| jith_ | mattoliverau: Thanks a lot... we are using this loopback partition instead of going for another added partition.. Am i right?? i have one more doubt.. we are creating 4 directories in /mntsdb1 as /mnt/sdb1/1 , /mnt/sdb1/2, /mnt/sdb1/3, and /mnt/sdb1/4 for replication purpose right??? and again why we are creating a softlink for the above 4 directories to /srv/1 /srv/2 /srv/3 and /srv/4 and... | 07:29 |
| jith_ | ...created another folder as node as /srv/num/node and using it?? | 07:29 |
| *** jordanP has joined #openstack-swift | 07:31 | |
| mattoliverau | jith_: you need to look in what's set up in the swift config files, the drives option is set to /srv/X/node. So the mounting and setting up of the SAIO XFS drive is done in mnt, and then the closer to real life view (or at least neater view) is placed in /srv. | 07:37 |
| mattoliverau | jith_: you must remember in the full cluster there will different nodes with there own drives, the SAIO is simulating multpile nodes, but all in one. So thats what /srv is trying to demonstrate, ie, keeping seperation in the nodes, even though they are stored all on the same disk | 07:38 |
| *** SkyRocknRoll has quit IRC | 08:00 | |
| ahale | a saio is probably one of the more complicated swift setups, kind of a pain as its also the entry point mostly | 08:04 |
| mattoliverau | ahale: yeah good point. SAIO, awesome to have, but if your new, but follow the set up instructions :P | 08:05 |
| ahale | and even if you're not new :) i regularly mess mine up | 08:05 |
| ahale | i'll be visiting your neck of the woods next week mattoliverau :) (though not your exact part I think - will be in sydney) | 08:08 |
| mattoliverau | ahale: cool! well it's closer then normal.. I'm in Melbourne. so still about 1000 kms away.. but still close. What brings you down here? If you came by Melb I'd have to make sure to take you out for a meal or something :) | 08:09 |
| ahale | Ah yep I thought you were there :) | 08:10 |
| ahale | just visiting my brother, meeting him in syd for a few days then spending a couple of weeks in nz south island | 08:11 |
| mattoliverau | Sydney has the main Rackspace office, but I think they're ashamed of me or something cause I never get to go visit :P | 08:11 |
| ahale | hehe, yeah I'm not visiting there :) | 08:11 |
| mattoliverau | ahale: cool :) NZ is beautiful too, so you'll have an awesome time! | 08:11 |
| ahale | I might drive by the DC and burn it down or something though.. no my favourite place | 08:11 |
| ahale | yep visited there before and really loved it, winters going to be a bit scary though | 08:12 |
| ahale | your winter seems warmer than londons summer though! | 08:13 |
| mattoliverau | lol, yeah, but NZ is Australia on training wheels, nothing will kill you.. well except the cold, mountains or extreme sports :P | 08:13 |
| mattoliverau | ahale: true dat | 08:13 |
| *** yuanz has joined #openstack-swift | 08:14 | |
| ahale | cold, mountains and extreme sports.. looks at plans to bungy in queenstown | 08:14 |
| *** chlong is now known as chlong_afk | 08:16 | |
| *** yuan has quit IRC | 08:17 | |
| mattoliverau | :) | 08:19 |
| *** jistr has joined #openstack-swift | 08:20 | |
| *** joeljwright has joined #openstack-swift | 08:20 | |
| mattoliverau | Time for dinner, night all! | 08:24 |
| *** haigang has joined #openstack-swift | 08:25 | |
| jith_ | mattoliverau: thanks a lot :) :) | 08:25 |
| *** haigang has quit IRC | 08:26 | |
| *** SkyRocknRoll has joined #openstack-swift | 08:26 | |
| ahale | gn | 08:30 |
| *** omame has joined #openstack-swift | 08:36 | |
| *** acoles_away is now known as acoles | 08:37 | |
| *** nadeem has quit IRC | 08:46 | |
| *** SkyRocknRoll has quit IRC | 08:46 | |
| *** bsdkurt has quit IRC | 08:50 | |
| *** SkyRocknRoll has joined #openstack-swift | 09:03 | |
| *** SkyRocknRoll has quit IRC | 09:03 | |
| *** SkyRocknRoll has joined #openstack-swift | 09:03 | |
| *** chlong_afk has quit IRC | 09:03 | |
| jith_ | i completed SAIO installation steps... what does the script in $HOME/swift/.unittests will check?? | 09:06 |
| *** bsdkurt has joined #openstack-swift | 09:08 | |
| zigo | Is there anyone here working on pyeclib? | 09:10 |
| zigo | I need help here for its packaging. | 09:10 |
| zigo | Version 1.0.7m has a wrong egg-info... | 09:10 |
| jith_ | hi all, after installation of swift when i start the swift using startmain. I received the following error: Exception: Could not bind to 127.0.0.1:6011 after trying for 30 seconds | 09:13 |
| *** haomaiwang has quit IRC | 09:17 | |
| *** aix has joined #openstack-swift | 09:22 | |
| ho | zigo: I think tsg_ works for pyeclib but i couldn't find his name in the list. | 09:23 |
| zigo | Thanks. | 09:23 |
| zigo | That's Tushar Gohad, but indeed, he's not here... | 09:24 |
| ho | zigo: yeah, i heard that begining of this week he was in japan | 09:26 |
| zigo | Ah, thanks. | 09:27 |
| *** acoles is now known as acoles_away | 09:28 | |
| zigo | So maybe end of the day for him already. | 09:28 |
| ho | jith_: to check the process of swift (1) execute resetswift (2) check the process of swift with ps command | 09:28 |
| jith_ | thanks ho.. when i execute startmain.. i got proxy-server already started... | 09:35 |
| jith_ | container-server running (29779 - /etc/swift/container-server/2.conf) | 09:35 |
| jith_ | container-server running (29780 - /etc/swift/container-server/3.conf) | 09:35 |
| jith_ | container-server running (29781 - /etc/swift/container-server/4.conf) | 09:35 |
| jith_ | container-server already started... | 09:35 |
| jith_ | container-server/1.conf is missing.. it is not started or with some error??? i m new to this | 09:36 |
| ho | jith_: i think you stop the processes with resetswift which is stored in the same place with startmain | 09:38 |
| jith_ | ho: yes i did... again when i execute startmain i got the following error: error: Exception: Could not bind to 127.0.0.1:6011 after trying for 30 seconds | 09:41 |
| jith_ | in ps it is running but when i do netstat -antulp | grep 6011... the output is | 09:42 |
| jith_ | (Not all processes could be identified, non-owned process info | 09:42 |
| jith_ | will not be shown, you would have to be root to see it all.) | 09:42 |
| jith_ | tcp 0 0 127.0.0.1:6011 0.0.0.0:* LISTEN - | 09:42 |
| jith_ | tcp6 0 0 ::1:6011 :::* LISTEN - | 09:42 |
| ho | jith_: do you make sure there is no swift related processes after resetswift? | 09:43 |
| jith_ | ho: yes.. but when i execute startmain.. and when i see the 6011 port using netstat.. it seems two services are running | 09:44 |
| ho | jith_: under container-server directory I think you have 1.conf, 2.conf, 3.conf and 4.conf. could you please check bind_port line in the config files whether there is no duplication? | 09:51 |
| jith_ | ho: sure | 09:52 |
| *** aix has quit IRC | 10:00 | |
| jith_ | ho: when i do lsof -i 6011.. i found like sshd service is running over there | 10:04 |
| jith_ | can i change the port no of tht particular container's conf file... | 10:04 |
| jith_ | after editing the port should i execute any other command to set it? | 10:06 |
| ho | jith_: i see and sorry for mis-leading. you already found out it with output of netstat. you can change it but you need to change ring files setting. is it possible to change sshd port? | 10:10 |
| ho | jith_: if you change the port in the config files, you need to change remakerings. | 10:12 |
| ho | acoles: thanks for the comment. I couldn't find Sam's attempt "IIRC he attempted to improve keystonemiddleware config to be more compatibel with paste deploy ini file". could you please teach me where it is? | 10:16 |
| jith_ | ho: thanks.. ya i edited in remakerings... so i have to execute remakerings and then i can start the swift... is it so?? | 10:18 |
| ho | jith_: yeah i think so. | 10:19 |
| *** aix has joined #openstack-swift | 10:28 | |
| *** dmorita has quit IRC | 10:35 | |
| jith_ | ho: thanks for ur guidance.. it worked.. how to connect this swift with running devstack machine...? | 10:41 |
| ho | jith_: Congratulation! (sorry for mis-leading even if you already found the duplication usage of port) | 10:43 |
| ho | jith_: from step 11 in http://docs.openstack.org/developer/swift/development_saio.html | 10:43 |
| jith_ | ho: no issues... thanks really for guiding.. i m a beginner | 10:50 |
| ho | jith_: thanks! have fun with swift :) | 10:54 |
| ho | acoles: The patch 190370 is adding info for keystone access log only so I think we don't need to hurry about landing. If there is no reason for hurry, we better to make a discussion for a way to support oslo config in swift first because this patch might introduce a backward compatibility issue on swift. | 10:54 |
| patchbot | ho: https://review.openstack.org/#/c/190370/ | 10:54 |
| ho | all, have a nice weekend! | 10:54 |
| jith_ | ho: how to increase the storage capacity? i initially configured with 20GB and if i need to make it as 40.. is it possible to create a same loopbackdevice and i can mount it inside the current swift directory.. or someother method to expand the GB | 10:55 |
| ho | jith_: add devices on ring. you need to add devices in remakerings | 10:56 |
| *** ho has quit IRC | 10:56 | |
| *** fifieldt has quit IRC | 10:58 | |
| jith_ | hi all any command to check the status of servers like resetting the swift using resetswift? | 10:59 |
| *** SkyRocknRoll has quit IRC | 11:35 | |
| *** km has quit IRC | 12:11 | |
| hurricanerix_ | jith_: Are you using a SAIO? You could try "swift-init all status". | 12:39 |
| *** NEW has joined #openstack-swift | 12:44 | |
| NEW | could anyone suggest which would be the best project to start with ..' | 12:44 |
| NEW | i want to be a core contributor and start in a growing project not already stable one, could anyone suggest some suggestion | 12:45 |
| *** bill_az has joined #openstack-swift | 12:49 | |
| *** NM has joined #openstack-swift | 12:53 | |
| mandarine | NEW: I've heard Gnocchi is quite a new and outstanding project ;) | 12:56 |
| NEW | I have never heard of it | 12:57 |
| NEW | what is it about | 12:57 |
| mandarine | Graphs and monitoring | 12:57 |
| mandarine | It's a "Time series database" thingy | 12:57 |
| NEW | anything related to connections less to data bases ? | 12:58 |
| mandarine | Nope | 12:58 |
| mandarine | Swifties, I've got another problem: I have got three servers (10CPU, 8GB RAM, 8x2TB HDD). I've got swift running on them with one replica. One customer is hitting on it and we get tons of load on disks and CPU from the object-server processes. Do you have any idea on what I'm missing? | 13:02 |
| mandarine | The thing works but cannot handle 1GBps traffic? I trust swift and I highly doubt that it's not able to do that ^^ | 13:03 |
| *** annegentle has joined #openstack-swift | 13:04 | |
| *** haypo has joined #openstack-swift | 13:09 | |
| *** kei_yama has quit IRC | 13:14 | |
| *** acampbell has joined #openstack-swift | 13:15 | |
| *** RackerShagz has joined #openstack-swift | 13:16 | |
| *** wbhuber has joined #openstack-swift | 13:16 | |
| *** petertr7_away is now known as petertr7 | 13:20 | |
| *** proteusguy has quit IRC | 13:28 | |
| mandarine | About my previous asking ... a reboot did the trick. I really hate this kind of things. I'll try to investigate it more when it reappears >_< | 13:29 |
| *** amoturi has joined #openstack-swift | 13:31 | |
| *** RackerShagz has quit IRC | 13:36 | |
| *** RackerShagz has joined #openstack-swift | 13:40 | |
| *** haypo has left #openstack-swift | 13:44 | |
| *** bill_az has quit IRC | 13:44 | |
| *** RackerShagz has quit IRC | 13:50 | |
| *** acoles_away is now known as acoles | 13:51 | |
| *** RackerShagz has joined #openstack-swift | 13:51 | |
| acoles | ho: ping | 13:52 |
| *** annegentle has quit IRC | 14:00 | |
| *** ChanServ sets mode: +v tdasilva | 14:06 | |
| *** annegentle has joined #openstack-swift | 14:07 | |
| *** wbhuber has quit IRC | 14:08 | |
| acoles | ho: nm i guess i missed your timezone. i'm not sure how we would be making a backwards compatibility problem by adding one more option to the sample conf? | 14:11 |
| acoles | ho: and we can't stop people adding the option to their actual conf files. | 14:12 |
| *** annegentle has quit IRC | 14:20 | |
| *** NM has quit IRC | 14:21 | |
| *** jrichli has joined #openstack-swift | 14:30 | |
| *** wbhuber has joined #openstack-swift | 14:38 | |
| *** remix_tj has joined #openstack-swift | 14:38 | |
| jith_ | hurricanerix_.. yes | 14:38 |
| jith_ | thanks | 14:38 |
| hurricanerix_ | jith_: np | 14:39 |
| *** SkyRocknRoll has joined #openstack-swift | 14:45 | |
| *** minwoob has joined #openstack-swift | 14:50 | |
| *** geaaru has joined #openstack-swift | 14:50 | |
| *** annegentle has joined #openstack-swift | 14:58 | |
| *** acoles is now known as acoles_away | 15:19 | |
| *** ppai has quit IRC | 15:19 | |
| jith_ | hi all ,if i need to add my standalone swift VM to another devstack VM.. swift endpoint will run in keystone only?? where i have to mention my swift details?? i did SAIO setup | 15:28 |
| jith_ | add in the sense, i need to configure swift device to the running devstack | 15:28 |
| *** annegentle has quit IRC | 15:28 | |
| *** annegentle has joined #openstack-swift | 15:29 | |
| *** petertr7 is now known as petertr7_away | 15:34 | |
| *** nadeem has joined #openstack-swift | 15:34 | |
| *** petertr7_away is now known as petertr7 | 15:38 | |
| *** amoturi has quit IRC | 15:44 | |
| *** haomaiwang has joined #openstack-swift | 15:44 | |
| *** zhill_desktop has joined #openstack-swift | 15:46 | |
| *** ptb has joined #openstack-swift | 15:46 | |
| *** kutija has joined #openstack-swift | 15:51 | |
| *** kutija_ has quit IRC | 15:55 | |
| *** proteusguy has joined #openstack-swift | 16:01 | |
| *** nadeem has quit IRC | 16:03 | |
| *** bill_az has joined #openstack-swift | 16:11 | |
| *** ptb has quit IRC | 16:12 | |
| *** nadeem has joined #openstack-swift | 16:17 | |
| *** diazjf has joined #openstack-swift | 16:22 | |
| *** gyee_ has joined #openstack-swift | 16:28 | |
| *** RackerShagz has quit IRC | 16:31 | |
| *** petertr7 is now known as petertr7_away | 16:35 | |
| *** NEW has quit IRC | 16:37 | |
| *** janonymous_ has joined #openstack-swift | 16:38 | |
| notmyname | good morning | 16:44 |
| notmyname | peluse: can you please ping me in here when your ec patch hits gerrit? just in case I miss it otherwise | 16:44 |
| notmyname | hurricanerix_: will you be able to give a status update on hummingbird in next week's meeting? (or redbo or dfg_ or glange) | 16:45 |
| *** ptb has joined #openstack-swift | 16:47 | |
| *** zaitcev has joined #openstack-swift | 16:47 | |
| *** ChanServ sets mode: +v zaitcev | 16:47 | |
| hurricanerix_ | notmyname: I probably would not be the best person for that, but if nobody else wants to do it, then I can. | 16:48 |
| *** ptb has quit IRC | 16:48 | |
| notmyname | hurricanerix_: thanks | 16:49 |
| *** joeljwright has quit IRC | 16:49 | |
| *** jordanP has quit IRC | 16:49 | |
| *** RackerShagz has joined #openstack-swift | 16:50 | |
| *** RackerSh_ has joined #openstack-swift | 16:54 | |
| *** RackerShagz has quit IRC | 16:55 | |
| *** jith_ has quit IRC | 16:55 | |
| *** RackerSh_ has quit IRC | 16:59 | |
| *** RackerShagz has joined #openstack-swift | 16:59 | |
| clayg | mandarine: did you say *one* replica? like you mean one replica per server or like raid under there? I think swift does a bunch of work that only makes sense in a durable storage layer - if you like the swift api but want to use it more like a volitole cache/non-primary storage/staging area we probably need to do some work to make that use case work well for you :\ | 17:02 |
| *** RackerSh_ has joined #openstack-swift | 17:03 | |
| *** RackerShagz has quit IRC | 17:05 | |
| *** annegentle has quit IRC | 17:05 | |
| *** RackerSh_ has quit IRC | 17:07 | |
| *** wbhuber has quit IRC | 17:15 | |
| *** breitz has quit IRC | 17:16 | |
| *** gyee_ has quit IRC | 17:23 | |
| *** RackerShagz has joined #openstack-swift | 17:29 | |
| *** occupant has quit IRC | 17:30 | |
| *** RackerShagz has quit IRC | 17:31 | |
| *** RackerShagz has joined #openstack-swift | 17:31 | |
| *** petertr7_away is now known as petertr7 | 17:33 | |
| *** occupant has joined #openstack-swift | 17:33 | |
| *** wbhuber has joined #openstack-swift | 17:46 | |
| *** ptb has joined #openstack-swift | 17:47 | |
| *** jistr has quit IRC | 17:48 | |
| *** ptb has quit IRC | 17:55 | |
| *** gyee_ has joined #openstack-swift | 17:57 | |
| *** annegentle has joined #openstack-swift | 17:58 | |
| *** nadeem has quit IRC | 18:01 | |
| mandarine | clayg: I meant two. Sorry about it ;-) | 18:01 |
| mandarine | One "data" and one "copy of the data" in my mind. | 18:01 |
| *** nadeem has joined #openstack-swift | 18:01 | |
| *** fthiagogv has joined #openstack-swift | 18:03 | |
| *** annegentle has quit IRC | 18:05 | |
| *** ptb has joined #openstack-swift | 18:06 | |
| *** acampbell has quit IRC | 18:10 | |
| openstackgerrit | janonymous proposed openstack/swift: Fixed Formatting Error in Swift -Form Post middleware section. https://review.openstack.org/191171 | 18:13 |
| *** SkyRocknRoll has quit IRC | 18:13 | |
| *** nadeem has quit IRC | 18:19 | |
| *** amoturi has joined #openstack-swift | 18:19 | |
| *** aix has quit IRC | 18:22 | |
| *** nadeem has joined #openstack-swift | 18:27 | |
| lcurtis | hello all. question about handoff nodes. If say I had a server go down and data was writeen to handoff node, does that data get rsync'd back to original node when it comes back online? | 18:30 |
| *** annegentle has joined #openstack-swift | 18:30 | |
| clayg | mandarine: right on man ;) | 18:34 |
| clayg | lcurtis: sure does | 18:34 |
| lcurtis | clayg awesome...thnks! | 18:35 |
| lcurtis | am i correct in stating that data from a failed node will not be rereplicated until weight of failed node devices is set to zero? | 18:36 |
| notmyname | lcurtis: node == server? | 18:37 |
| lcurtis | notmyname yes | 18:37 |
| notmyname | lcurtis: yup | 18:37 |
| lcurtis | thank you so much | 18:37 |
| lcurtis | starting to get the questions | 18:37 |
| lcurtis | want to make sure i am stating them accurately | 18:37 |
| lcurtis | the answers that is | 18:37 |
| notmyname | lcurtis: basically, if a failure is because it could connect to the server but not read/write the drive (ie a 507 response), data will be replicated | 18:37 |
| notmyname | lcurtis: and if it cannon connect at all, it won't be replicated until it's removed from the ring | 18:38 |
| lcurtis | interesting...so if a drive fails...data does indeed start replicating | 18:39 |
| lcurtis | ? | 18:39 |
| lcurtis | sp a 507 would cause data to go to handoff servers? | 18:40 |
| lcurtis | so | 18:40 |
| notmyname | correct | 18:41 |
| lcurtis | brilliant...love it | 18:41 |
| lcurtis | thank you | 18:41 |
| mandarine | clayg: as said previously, doing a reboot seems to have solved the things. I feel a "CPU leak" in this ;) | 18:43 |
| openstackgerrit | John Dickinson proposed openstack/swift: do container listing updates in another (green)thread https://review.openstack.org/189080 | 18:43 |
| *** nadeem has quit IRC | 18:44 | |
| mandarine | I used to hate rebooting production servers, but I just realise : it still works with two out of three servers and that is why we did that ... | 18:44 |
| clayg | works like a champ | 18:46 |
| mandarine | Yup. I'm not used to that ;) | 18:48 |
| *** annegentle has quit IRC | 18:50 | |
| *** acampbell has joined #openstack-swift | 18:52 | |
| timburke | wow. i have a change so old, it's even fallen off of the "Older open patches" section of the review dashboard. anyone mind taking a look at patch 161972? | 18:52 |
| patchbot | timburke: https://review.openstack.org/#/c/161972/ | 18:52 |
| *** annegentle has joined #openstack-swift | 18:54 | |
| *** ronenkat has joined #openstack-swift | 18:55 | |
| *** petertr7 is now known as petertr7_away | 19:05 | |
| *** petertr7_away is now known as petertr7 | 19:05 | |
| *** mfalatic has quit IRC | 19:07 | |
| *** acampbell has quit IRC | 19:08 | |
| *** ronenkat has quit IRC | 19:09 | |
| *** dontalton has joined #openstack-swift | 19:09 | |
| *** ptb has quit IRC | 19:10 | |
| *** RackerShagz has quit IRC | 19:17 | |
| *** RackerShagz has joined #openstack-swift | 19:18 | |
| *** acampbell has joined #openstack-swift | 19:19 | |
| *** RackerShagz has quit IRC | 19:20 | |
| *** vinsh_ has joined #openstack-swift | 19:21 | |
| *** RackerShagz has joined #openstack-swift | 19:21 | |
| *** RackerSh_ has joined #openstack-swift | 19:23 | |
| *** annegentle has quit IRC | 19:24 | |
| *** RackerSh_ has quit IRC | 19:24 | |
| *** vinsh has quit IRC | 19:24 | |
| *** RackerSh_ has joined #openstack-swift | 19:24 | |
| *** RackerShagz has quit IRC | 19:25 | |
| *** ptb has joined #openstack-swift | 19:29 | |
| *** wer has quit IRC | 19:33 | |
| *** wer has joined #openstack-swift | 19:45 | |
| *** blmartin has joined #openstack-swift | 19:46 | |
| *** dontalton has quit IRC | 19:52 | |
| *** ronenkat has joined #openstack-swift | 19:52 | |
| *** acampbell has quit IRC | 19:55 | |
| openstackgerrit | Darrell Bishop proposed openstack/swift: Allow 1+ object-servers-per-disk deployment https://review.openstack.org/184189 | 19:57 |
| swifterdarrell | clayg: ^^^^^^^^^^^^^^^^^^^ | 19:58 |
| *** proteusguy has quit IRC | 19:58 | |
| swifterdarrell | zaitcev: mattoliverau: notmyname: torgomatic: ^^^^^^^^^^ | 19:59 |
| zaitcev | swifterdarrell: Thanks | 20:00 |
| swifterdarrell | zaitcev: I hope you like how the swift.common.wsgi crap ended up :) | 20:01 |
| *** ronenkat has quit IRC | 20:01 | |
| *** petertr7 is now known as petertr7_away | 20:01 | |
| *** ptb has quit IRC | 20:04 | |
| *** petertr7_away is now known as petertr7 | 20:04 | |
| *** nadeem has joined #openstack-swift | 20:07 | |
| *** blmartin has quit IRC | 20:07 | |
| *** ian_brown has quit IRC | 20:07 | |
| *** swat30 has quit IRC | 20:07 | |
| *** ahale has quit IRC | 20:07 | |
| *** patchbot has quit IRC | 20:07 | |
| *** chmouel has quit IRC | 20:07 | |
| *** ndk has quit IRC | 20:07 | |
| *** blmartin_ has joined #openstack-swift | 20:07 | |
| *** ndk has joined #openstack-swift | 20:08 | |
| *** patchbot has joined #openstack-swift | 20:08 | |
| *** swat30 has joined #openstack-swift | 20:08 | |
| *** ahale has joined #openstack-swift | 20:10 | |
| *** ian_brown has joined #openstack-swift | 20:10 | |
| *** proteusguy has joined #openstack-swift | 20:10 | |
| *** EmilienM is now known as EmilienM|afk | 20:10 | |
| *** chmouel has joined #openstack-swift | 20:12 | |
| *** RackerShagz has joined #openstack-swift | 20:21 | |
| *** RackerSh_ has quit IRC | 20:22 | |
| clayg | swifterdarrell: I've really been loving your "this crap is all better in the next version" and "I fixed all this shit" comments in this change series :) | 20:22 |
| swifterdarrell | clayg: what can I say? the shit keeps getting better | 20:23 |
| *** amoturi has quit IRC | 20:25 | |
| clayg | swifterdarrell: heh, yeah I guess you have to drop that assert desired == actual :\ | 20:26 |
| swifterdarrell | clayg: yup... | 20:26 |
| *** greghaynes has quit IRC | 20:27 | |
| *** annegentle has joined #openstack-swift | 20:29 | |
| clayg | redbo: torgomatic: acoles_away: cschwede: ho: mattoliverau: jrichli: peluse: notmyname: dfg_: hurricanerix_: chmouel: portante: briancline: if anyone was waiting for things to settle down before reviewing the object-server-per-disk change - now is the time to get your hands dirty -> https://review.openstack.org/#/c/184189/ | 20:32 |
| portante | hmmm tempting | 20:34 |
| clayg | ... acctually if someone had a machine not doing anything I think most of my probtests during review were on a server-per-disk configuration | 20:35 |
| clayg | so if someone wanted to invest some cpu cycles w/o too much personal effort it'd be nice to grab your current swift-all-in-one setup (that currently passes probetests) and checkout that change and run probetests | 20:36 |
| clayg | *personally* i don't care - everyone should move to 127.0.0.1-4 instead of the port based tricks on their swift-all-in-ones anyway - but someone might bitch at me if probetests break for them | 20:36 |
| *** RackerShagz has quit IRC | 20:38 | |
| swifterdarrell | FWIW, I tried to keep the passing in "old style" (127.0.0.1x4 w/unique ports), "new" (127.0.0.1,2,3,4 w/unique ports), "new+servers_per_port" (127.0.0.1,2,3,4 w/ports unique within "server" only) | 20:38 |
| swifterdarrell | What probably won't work is 127.0.0.1x4 + servers_per_port > 0 | 20:39 |
| *** ozialien has joined #openstack-swift | 20:44 | |
| hurricanerix_ | clayg: I'll add that review to my Monday todo list. | 20:46 |
| *** annegentle has quit IRC | 20:48 | |
| *** tmckay has joined #openstack-swift | 20:58 | |
| clayg | hurricanerix_: baller! | 21:00 |
| tmckay | hi folks. playing with a multi-node setup with packstack (Juno). I would like to deploy multiple swift storage nodes but I'm not sure how to do that. CONFIG_SWIFT_STORAGE_HOSTS is deprecated, there is only CONFIG_SWIFT_STORAGES in the answerfile. Do you know of any guides out there for packstack? | 21:01 |
| clayg | tmckay: what is packstack? | 21:02 |
| clayg | tmckay: is there an #openstack-packstack channel? | 21:02 |
| tmckay | there may be. packstack is a deployment script you can use for RDO | 21:03 |
| clayg | https://wiki.openstack.org/wiki/Packstack#IRC says it's #packsatck-dev | 21:03 |
| *** blmartin_ has quit IRC | 21:03 | |
| clayg | RDO! zaitcev will know everything about it then! | 21:03 |
| tmckay | could be other things besides RDO I suppose, but that's what I | 21:04 |
| tmckay | I'm using. | 21:04 |
| tmckay | clayg, thanks, I saw the ref to packstack-dev, I thought I would try here first :) | 21:04 |
| clayg | sorry - i just know anything about that tool | 21:05 |
| tmckay | clayg, np, until recently I have only used devstack ;-) Trying to broaden my horizons (haha) | 21:06 |
| clayg | tmckay: nice! | 21:06 |
| tmckay | clayng, fought with neutron for days | 21:06 |
| clayg | tmckay: do you mainly develop on openstack - what's your intrest with swift? | 21:06 |
| tmckay | finally have multi-node compute with routes in and out of VMs over neutron | 21:06 |
| tmckay | clayg, I am a Sahara core team member, and I work with a bunch of guys who do Sahara/Trove. Trying to set up a test lab for us. | 21:07 |
| *** petertr7 is now known as petertr7_away | 21:07 | |
| tmckay | Sahara does read/write with Swift (as an option) | 21:07 |
| clayg | tmckay: wow! awesome! | 21:07 |
| tmckay | big data processing (data sets) | 21:08 |
| clayg | tmckay: yeah for sure! does it have some trick to make it easy to export a data/result set from hdfs to swift? | 21:08 |
| clayg | tmckay: we deploy a bunch of swift and when people have an hdfs cluster there always looking to dump some of the old stuff thye're using less but don't want to get rid of off there hdfs clusters | 21:09 |
| tmckay | yes, actually, we have a custom hdfs jar that extends hadoop to be able to write to swift paths | 21:09 |
| tmckay | so, the hadoop jobs read/write directly to the swift service in the stack | 21:09 |
| tmckay | the code was merged in Hadoop, but we have found that we move faster than Apache so Sahara carries its own patch currently and add it in with diskimage builder | 21:10 |
| notmyname | that's cool | 21:10 |
| tmckay | yeah, works well | 21:11 |
| *** wbhuber has quit IRC | 21:12 | |
| *** wbhuber has joined #openstack-swift | 21:12 | |
| *** tmckay has left #openstack-swift | 21:18 | |
| *** annegentle has joined #openstack-swift | 21:19 | |
| *** ozialien has quit IRC | 21:26 | |
| *** gyee_ has quit IRC | 21:33 | |
| *** fthiagogv has quit IRC | 21:42 | |
| -openstackstatus- NOTICE: Gerrit will be offline for project renames between 22:00 and 22:30 UTC | 21:44 | |
| *** ChanServ changes topic to "Gerrit will be offline for project renames between 22:00 and 22:30 UTC" | 21:44 | |
| *** nadeem has quit IRC | 21:51 | |
| *** nadeem has joined #openstack-swift | 21:52 | |
| *** jrichli has quit IRC | 21:54 | |
| *** jrichli has joined #openstack-swift | 21:54 | |
| *** jrichli has quit IRC | 21:54 | |
| *** diazjf has left #openstack-swift | 21:57 | |
| *** minwoob has quit IRC | 22:03 | |
| -openstackstatus- NOTICE: Gerrit is offline for project renames. ETA 20:30 | 22:05 | |
| *** ChanServ changes topic to "Gerrit is offline for project renames. ETA 20:30" | 22:05 | |
| -openstackstatus- NOTICE: Gerrit is offline for project renames. ETA 22:40 | 22:10 | |
| *** ChanServ changes topic to "Gerrit is offline for project renames. ETA 22:40" | 22:10 | |
| *** wer_ has joined #openstack-swift | 22:16 | |
| *** wer has quit IRC | 22:17 | |
| *** amoturi has joined #openstack-swift | 22:24 | |
| *** annegentle has quit IRC | 22:34 | |
| *** annegentle has joined #openstack-swift | 22:34 | |
| *** sam9987 has quit IRC | 22:36 | |
| *** lcurtis has quit IRC | 22:36 | |
| *** geaaru has quit IRC | 22:40 | |
| peluse | notmyname, clayg: I'm not going to have time to submit that Ec recon patch today, its working great though on the cluster I just need to fixup existing test code and add a bit. Won't take long though, just not time today. Will knock it out Sat.... | 22:43 |
| notmyname | ok | 22:43 |
| peluse | BTW, quick update on the performance work on the 14 node clluster. All is running well, we are filling to 50% this weekend and will start real runs on Mon | 22:44 |
| *** ChanServ changes topic to "Gerrit is offline for project renames. ETA 20:30" | 22:44 | |
| -openstackstatus- NOTICE: Gerrit is back online. Zuul reconfiguration for renamed projects is still in progress, ETA 23:30. | 22:44 | |
| *** zaitcev has quit IRC | 22:44 | |
| *** zaitcev has joined #openstack-swift | 22:45 | |
| *** ChanServ sets mode: +v zaitcev | 22:45 | |
| *** annegentle has quit IRC | 22:45 | |
| *** marcusvrn_ has joined #openstack-swift | 22:47 | |
| *** ChanServ changes topic to "Review Dashboard: http://goo.gl/8IUcKl | Summary Dashboard: http://goo.gl/qHus5v | Summit scheduling: https://etherpad.openstack.org/p/liberty-swift-summit-topics | Logs: http://eavesdrop.openstack.org/irclogs/%23openstack-swift/" | 22:50 | |
| *** marcusvrn_ is now known as marcusvrn | 22:52 | |
| *** amoturi has quit IRC | 23:06 | |
| notmyname | I got the black swift keycaps for my keyboard. they look awesome! | 23:09 |
| *** zhill_desktop has quit IRC | 23:11 | |
| notmyname | trying to get the photo synced to my computer so I can share it | 23:11 |
| clayg | for all of you out there in internetland i've already seen them IRL and they do look pretty swanky | 23:12 |
| notmyname | http://d.not.mn/black_swift_keycaps.jpg | 23:17 |
| *** openstackgerrit has quit IRC | 23:22 | |
| *** openstackgerrit has joined #openstack-swift | 23:22 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
