*** thumpba_ has joined #openstack-swift | 00:03 | |
*** thumpba has quit IRC | 00:05 | |
*** Jeffrey4l has joined #openstack-swift | 00:16 | |
csmart | notmyname: did you get to see/play with any Intel Joule dev kits at IDF? | 00:41 |
---|---|---|
*** Suyi has quit IRC | 00:44 | |
*** jamielennox is now known as jamielennox|away | 00:48 | |
notmyname | csmart: I did not. but there were big signs all over | 00:52 |
notmyname | csmart: I'll be going back tomorrow, though | 00:52 |
csmart | kk, just curious | 00:52 |
*** tqtran has quit IRC | 00:54 | |
*** m_kazuhiro has joined #openstack-swift | 00:54 | |
notmyname | csmart: it looks interesting. until you realize it's literally 10x the cost of a raspberry pi. so how do they justify marketing it against the pi? | 00:57 |
notmyname | csmart: but if you want me to buy you one and mail it to you, I can | 00:58 |
csmart | notmyname: exactly. I'm not sure what intel's plan is in the IoT space... | 00:58 |
csmart | notmyname: thanks, appreciate the offer, but I think it would just become one more gadget on my shelf :-S | 00:59 |
notmyname | csmart: intel + IoT (honestly, everybody + IoT) reminds me of https://www.youtube.com/watch?v=9ntPxdWAWq8 | 00:59 |
*** adu has quit IRC | 01:00 | |
*** jamielennox|away is now known as jamielennox | 01:04 | |
*** cdelatte has joined #openstack-swift | 01:06 | |
clayg | do we have an interable to decompressing file like? | 01:08 |
clayg | like convert a log.gz in swift to a swiftclient get_object resp_chunk_size iterable to a readable suitable `for line in ~MAGIC~` | 01:08 |
clayg | because I don't want to write that again :'( | 01:09 |
*** airdisa has quit IRC | 01:10 | |
clayg | oh internal client you wonderful thing | 01:10 |
*** klrmn has quit IRC | 01:11 | |
clayg | totally un-used in-tree method "iter_object_lines" -> "Returns an iterator of object lines from an uncompressed or compressed text object" | 01:11 |
notmyname | clayg: COmpressingFileReader or something like that | 01:13 |
clayg | notmyname: no that's the wrong way - surprissingly it was the first thing I thought of | 01:14 |
notmyname | if it's no longer in tree, I've got one that at least used to be for all the log proces....oh ok | 01:14 |
clayg | glange_: thanks for iter_object_lines in InternalClient! | 01:14 |
*** chrichip has joined #openstack-swift | 01:19 | |
csmart | notmyname: lol gold | 01:20 |
csmart | "crowdsourcing is something we are having" | 01:20 |
csmart | must. not. click. more. the onion videos. | 01:22 |
openstackgerrit | zheng yin proposed openstack/swift: Remove unnecessary variable https://review.openstack.org/356402 | 01:34 |
zhengyin | timburke: thank you for your comment with patch 356402. I add more commit message :) | 01:41 |
patchbot | zhengyin: https://review.openstack.org/#/c/356402/ - swift - Remove unnecessary variable | 01:41 |
*** nadeem has joined #openstack-swift | 01:47 | |
*** delattec has joined #openstack-swift | 01:47 | |
*** cdelatte has quit IRC | 01:49 | |
*** tqtran has joined #openstack-swift | 01:51 | |
*** thumpba_ has quit IRC | 01:52 | |
*** tqtran has quit IRC | 01:56 | |
*** bill_az has quit IRC | 01:58 | |
*** nadeem has quit IRC | 02:05 | |
*** delattec has quit IRC | 02:09 | |
*** chrichip has quit IRC | 02:13 | |
*** delattec has joined #openstack-swift | 02:14 | |
*** chrichip has joined #openstack-swift | 02:15 | |
*** diogogmt has joined #openstack-swift | 02:17 | |
*** diogogmt has quit IRC | 02:18 | |
*** thumpba has joined #openstack-swift | 02:23 | |
openstackgerrit | Tuan proposed openstack/swift: Remove except: using except Exception: https://review.openstack.org/356833 | 02:25 |
*** klrmn has joined #openstack-swift | 02:28 | |
*** dmorita has quit IRC | 02:28 | |
*** psachin has quit IRC | 02:40 | |
*** jamielennox is now known as jamielennox|away | 02:42 | |
zhengyin | notmyname: hi notmyname, do you have time to help me review a patch 354048? thanks :) | 02:45 |
patchbot | zhengyin: https://review.openstack.org/#/c/354048/ - swift - fix AccountReaper and unit test | 02:45 |
*** chrichip has quit IRC | 02:45 | |
*** chrichip has joined #openstack-swift | 02:47 | |
openstackgerrit | Janie Richling proposed openstack/swift: Add test for POST to DLO manifest file https://review.openstack.org/347545 | 03:06 |
openstackgerrit | Janie Richling proposed openstack/swift: Allow DLO manifest to be updated when using post-as-copy https://review.openstack.org/356239 | 03:06 |
notmyname | zhengyin: good evening | 03:06 |
*** delattec has quit IRC | 03:06 | |
*** jamielennox|away is now known as jamielennox | 03:07 | |
*** asettle has joined #openstack-swift | 03:08 | |
zhengyin | notmyname: good evening | 03:08 |
openstackgerrit | Janie Richling proposed openstack/swift: Add test for POST to DLO manifest file https://review.openstack.org/347545 | 03:09 |
jrichli | acoles: ^^ that is reverting back. sorry, rebasing issues. | 03:10 |
notmyname | zhengyin: I looked over it. and I have a few suggestions. would you like to chat about it just a bit? | 03:15 |
*** thumpba has quit IRC | 03:15 | |
*** asettle has quit IRC | 03:16 | |
zhengyin | notmyname: ok | 03:17 |
notmyname | zhengyin: ok. 2 things | 03:18 |
notmyname | zhengyin: first, let's look at the test_reset_stats() method | 03:18 |
notmyname | zhengyin: what you have is good | 03:18 |
notmyname | zhengyin: it creates and instance, and the constructor is calling reset_stats. then you check that the attributes are reset | 03:19 |
notmyname | so that's good | 03:19 |
zhengyin | thank you | 03:19 |
notmyname | however, it's not super clear that it's actually resetting anything | 03:19 |
notmyname | hang on..looking for what I'm thinking about | 03:21 |
notmyname | zhengyin: ok, so for example, if *after( the current method you added some lines that actually set those instance attributes to something, then you called reset_stats, then you checked that they are all zero...that would be nice | 03:22 |
notmyname | zhengyin: that's the first thing. the 2nd is just some formatting on the commit message | 03:23 |
notmyname | zhengyin: but those are both minor issues | 03:23 |
notmyname | zhengyin: and I'm curious about your thoughts on that suggestion | 03:23 |
zhengyin | notmyname:minor issues? | 03:25 |
notmyname | yeah. not big changes that substantially change the good work you've done, but things that make future us more happy | 03:27 |
zhengyin | notmyname: thanks, I will continue to improve them | 03:29 |
zhengyin | notmyname: including commit message | 03:30 |
notmyname | zhengyin: this is what I mean https://gist.github.com/notmyname/54985aad314f46a015a14397adcfe08f | 03:31 |
notmyname | does that make sense? | 03:31 |
notmyname | (definitely check it for correctness) | 03:31 |
notmyname | zhengyin: and I'd make the commit message https://gist.github.com/notmyname/0231af45c669b744eefe8e3104673363 | 03:32 |
notmyname | zhengyin: I don't care if you use my code there or not. I just wanted to show an exampel of what I was thinking of | 03:33 |
zhengyin | notmyname: thanks, I will read and learn commit message :) | 03:33 |
zhengyin | notmyname:yes, I know | 03:35 |
openstackgerrit | Merged openstack/swift: Drop X-Auth-Token from all versioned_writes subrequests https://review.openstack.org/307929 | 03:37 |
zhengyin | notmyname: I read your code and I know your thanking of . thanks a lot :) | 03:38 |
notmyname | zhengyin: no. thank you. you're the one writing code for swift. :-) | 03:39 |
zhengyin | notmyname: :) | 03:41 |
*** furlongm has quit IRC | 03:43 | |
*** furlongm has joined #openstack-swift | 03:43 | |
*** links has joined #openstack-swift | 03:45 | |
notmyname | csmart: http://www.intel.com/content/www/us/en/software/intel-cache-acceleration-software-performance.html and specifically http://www.intel.com/content/www/us/en/software/cas-for-linux-product-brief.html gives a reasonable overview | 03:48 |
csmart | notmyname: will CAS need a userspace daemon, or will it be kernel /sys tweaking? | 03:48 |
csmart | notmyname: hehe thanks, good timing | 03:48 |
notmyname | csmart: and http://softwarestore.ispfulfillment.com/Store/Product.aspx?skupart=I24S02 shows it's clearly (currently) positioned away from the hobbiest/tinkerer demographic | 03:49 |
notmyname | csmart: just enough to chase away students. not enough to make any actual business care | 03:49 |
csmart | interesting "Intel® CAS software is licensed per SSD and is included with the Intel® SSD at time of purchase with one-year of included support." | 03:51 |
csmart | I often wondered why this sort of cache accelleration was not already in the kernel | 03:51 |
csmart | *acceleration | 03:52 |
notmyname | yeah. and I went by the booth to see the results and meet the people face-to-face (you know, read the blog). but I was surprised when he said they were planning on opening it | 03:52 |
*** tqtran has joined #openstack-swift | 03:52 | |
notmyname | makes sense, though, since software has never been intel's thing | 03:52 |
csmart | yeah, well if they want to distribute the kernel driver it needs to GPL'd and I think they've been pretty good at opening those bits, at least | 03:53 |
csmart | *distribute it pre-compiled against a kernel | 03:53 |
notmyname | yeah. I'm looking at some docs I have now | 03:53 |
notmyname | the kernel module is called "intelcas" | 03:54 |
csmart | it seems there will be something proprietary at userspace layer, presumably for licensing and also configuring/tweaking the loads (I assume this will be where much of the smarts will actuall exist) | 03:54 |
notmyname | casadm is the admin controller to configure it | 03:54 |
notmyname | no, AIUI, the smarts are in the kernel module, since it's actually inspecting the traffic and categorizing it | 03:55 |
csmart | ok, that's good | 03:55 |
notmyname | eg (and what really interests me) is that it can detect filesystem metadata operations and cache those and pass through everything else | 03:55 |
*** tqtran has quit IRC | 03:56 | |
notmyname | "The Intel® Cache Acceleration Software installs into the Linux* operating system as a kernel module. The nature of the integration provides a cache solution that is transparent to users, applications, and your existing storage infrastructure. No storage migration or application changes are required." | 03:57 |
csmart | ok, interesting - I'd have thought some configuration to tweak for a specific load would be necessary. kinda cool if that all happens in the kernel | 03:57 |
notmyname | yeah, the casadm tool is used to configure the cache devices and what it's fronting. and loading the config file that sets the cache eviction policy. and does some other management stuff | 03:58 |
csmart | k | 03:58 |
notmyname | Intel® Cache Acceleration Software now provides the ability to control caching of your data at a finer granularity than ever before. Intel® Cache Acceleration Software for Linux can analyze every IO on the fly to determine whether the requested block is filesystem metadata or data and, if it is data, the total size of file IO requested. Using this information the administrator can determine the best IO cla | 03:59 |
notmyname | ss configuration settings for the typical workload and can set which IO classes to cache and which not to cache, and set a priority level for each IO class. Then, when it becomes necessary to evict a cache line, the software will evict cache lines of the lowest available priority first (an improvement compared to traditional LRU eviction). | 03:59 |
csmart | looks like US$300 gets you a license for unlimited GB with 1 yr support (when paired with an intel ssd) - I'm kinda surprised that given it's that low that they bother charging for it. Why not give it away (but charge for support) and capture the market? | 03:59 |
notmyname | I just don't see how that's possible in userspace | 03:59 |
notmyname | yeah, IMO the price is either (1) product people not sure what they want to do or (2) eliminating costly "free" users that aren't actually helping their bottom line | 04:00 |
csmart | true, but those free users still bought an intel ssd.. | 04:00 |
notmyname | so you can actually use CAS with any sort of devices. you *could* have an HDD fron another HDD. that would be bad... | 04:00 |
notmyname | no, not really | 04:00 |
notmyname | yeah, this is a big-picture tactic for intel to puch more flash. but afaict no intel silicon is actually required | 04:01 |
csmart | oh, ok seems like you need like a 3700 series drive | 04:01 |
csmart | *I thought | 04:01 |
csmart | sweet | 04:01 |
*** bill_az has joined #openstack-swift | 04:03 | |
notmyname | just so happens that I've been testing it with the 2TB PCIe DC 3x00. they are very nice ;-) | 04:03 |
notmyname | I don't remember if it's the 35, 36, or 37 hunder series | 04:04 |
csmart | I just accidentally incorrectly bought an Intel 750 pcie 2.5" (for openstack dev), now I just need a U.2 to M.2 to PCIe converter converter and I'll be set :-S | 04:05 |
notmyname | heh | 04:06 |
csmart | shop had 1200GB (with picture of PCIe device), 800GB (with picture of PCIe device) and 400GB (no picture) versions next to each other - who knew that 400GB was a 2.5" drive and they didn't sell the 400GB PCIe version | 04:08 |
csmart | *PCIe card version | 04:08 |
notmyname | csmart: https://gist.github.com/notmyname/47bf1a6647107d359d577f664fff6681 | 04:15 |
*** Trixboxer has quit IRC | 04:15 | |
notmyname | csmart: oh! haha! it's on the public interwebz. so there. http://www.intel.com/content/dam/support/us/en/documents/software/data-center-software/CAS-Linux_AG_3.0_GA.pdf | 04:18 |
notmyname | csmart: that has a _lot_ of detail in it | 04:19 |
*** Trixboxer has joined #openstack-swift | 04:28 | |
*** dmorita has joined #openstack-swift | 04:29 | |
notmyname | csmart: "Swift architecture running on OpenPOWER vs Intel x86 CPUs." <-- from a proposed talk on running swift on OpenPOWER | 04:32 |
notmyname | csmart: you know anything about that? | 04:32 |
*** dmorita has quit IRC | 04:33 | |
notmyname | csmart: title is "The Open Hardware Alternative: Geo-Distributed Swift Deployments at Scale with OpenPOWER" | 04:35 |
zaitcev | At least it's not s390. | 04:35 |
notmyname | redbo: did you see this talk proposed? "Adventures in exploring containers : Using, Cloudfuse to get Swift as Docker volume" | 04:35 |
notmyname | redbo: I'm thinking of advocating for it just to see your reaction ;-) | 04:36 |
redbo | ha. I get emails from people trying to chain cloudfuse in all sorts of crazy ways. | 04:39 |
notmyname | redbo: I suggest now you just reply with "have you tried running it in a docker container?" | 04:39 |
*** chrichip has quit IRC | 04:53 | |
*** chrichip has joined #openstack-swift | 04:53 | |
openstackgerrit | Merged openstack/swift: Don't run probe tests if resetswift failed https://review.openstack.org/355687 | 04:54 |
openstackgerrit | Merged openstack/swift: Add rm to tox whitelist_externals https://review.openstack.org/356584 | 04:55 |
csmart | notmyname: cool thanks.. | 05:16 |
* csmart adds it to his reading list | 05:17 | |
csmart | notmyname: hmm.. no I don't know anything about that.. was it related to that IBM guy we talked about before? | 05:17 |
csmart | Tom Mathews? | 05:17 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 05:33 |
*** asettle has joined #openstack-swift | 05:35 | |
*** SkyRocknRoll has joined #openstack-swift | 05:37 | |
*** asettle has quit IRC | 05:39 | |
*** rcernin has joined #openstack-swift | 06:12 | |
*** takashi has joined #openstack-swift | 06:35 | |
*** bill_az has quit IRC | 06:49 | |
*** mingyu has joined #openstack-swift | 06:49 | |
*** zaitcev has quit IRC | 06:49 | |
*** pcaruana has joined #openstack-swift | 06:52 | |
*** tesseract- has joined #openstack-swift | 07:17 | |
kota_ | mahatic_: quick fix is https://gist.github.com/bloodeagle40234/3d05f8cd4c217fc62581163e3a8e3078 | 07:33 |
kota_ | mahatic_: for your pach https://review.openstack.org/#/c/346865/ | 07:33 |
patchbot | kota_: patch 346865 - swift - Delete old tombstones | 07:33 |
kota_ | oh, what? if I don't set the reclaim_age value there, no tests failed??? | 07:35 |
kota_ | am i missing something? | 07:35 |
mahatic_ | kota_: in a meeting. be back in 10 | 07:36 |
kota_ | k | 07:37 |
onovy | mattoliverau: !!!! :) | 07:43 |
*** klrmn has quit IRC | 07:45 | |
*** tqtran has joined #openstack-swift | 07:54 | |
*** tqtran has quit IRC | 07:59 | |
*** rledisez has joined #openstack-swift | 08:02 | |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/swift: [WIP] Testing latest u-c https://review.openstack.org/318441 | 08:10 |
*** mingyu has quit IRC | 08:23 | |
*** dosaboy_ has joined #openstack-swift | 08:23 | |
*** mingyu has joined #openstack-swift | 08:23 | |
*** dosaboy_ has quit IRC | 08:24 | |
*** dosaboy has quit IRC | 08:24 | |
*** kei_yama_ has joined #openstack-swift | 08:25 | |
*** kei_yama has quit IRC | 08:26 | |
*** dosaboy has joined #openstack-swift | 08:29 | |
openstackgerrit | Merged openstack/swift: Change schedule priority of daemon/server in config https://review.openstack.org/238799 | 08:34 |
*** asettle has joined #openstack-swift | 08:39 | |
onovy | tadaaaaaaaaa | 08:46 |
onovy | #323819 #323874 // anyone ( notmyname tdasilva kota_ ) this simple reviews pls? | 08:56 |
onovy | patch 323819 | 08:56 |
patchbot | onovy: https://review.openstack.org/#/c/323819/ - liberasurecode - Changed homepage to new one | 08:56 |
onovy | patch 323874 | 08:56 |
patchbot | onovy: https://review.openstack.org/#/c/323874/ - pyeclib - Change liberasure URL to new one | 08:56 |
openstackgerrit | Christian Schwede proposed openstack/swift: Add support to increase object ring partition power https://review.openstack.org/337297 | 08:57 |
openstackgerrit | Ondřej Nový proposed openstack/swift: Correctly check all return codes in SAIO tests https://review.openstack.org/357025 | 08:57 |
onovy | kota_: thanks | 08:59 |
kota_ | onovy: np, sorry slow reviews. | 08:59 |
kota_ | onovy: btw, do you maintain the debian package right? | 09:00 |
openstackgerrit | Merged openstack/liberasurecode: Changed homepage to new one https://review.openstack.org/323819 | 09:02 |
openstackgerrit | Merged openstack/pyeclib: Change liberasure URL to new one https://review.openstack.org/323874 | 09:03 |
*** ouchkernel has quit IRC | 09:04 | |
*** joeljwright has joined #openstack-swift | 09:05 | |
*** ChanServ sets mode: +v joeljwright | 09:05 | |
*** d0ugal has quit IRC | 09:08 | |
*** ouchkernel has joined #openstack-swift | 09:09 | |
*** d0ugal has joined #openstack-swift | 09:12 | |
*** d0ugal_ has joined #openstack-swift | 09:18 | |
*** d0ugal_ has quit IRC | 09:18 | |
*** d0ugal_ has joined #openstack-swift | 09:18 | |
*** d0ugal has quit IRC | 09:20 | |
*** d0ugal_ has quit IRC | 09:20 | |
*** d0ugal has joined #openstack-swift | 09:20 | |
*** mvk has quit IRC | 09:29 | |
*** acoles_ is now known as acoles | 09:34 | |
acoles | jrichli: looks like something went awry rebasing patch 347545? ping me when you are in and we can get it sorted | 09:40 |
patchbot | acoles: https://review.openstack.org/#/c/347545/ - swift - Add test for POST to DLO manifest file | 09:40 |
onovy | kota_: yep | 09:45 |
onovy | liberasure + pyeclib | 09:45 |
kota_ | onovy: good | 09:45 |
kota_ | onovy: so I wonder if you could update the liberasurecode to the newest release in the packages | 09:46 |
kota_ | onovy: liberasurecode is now 1.2.0 and the release includes some improvement which needed in the newest PyECLib | 09:46 |
*** jordanP has joined #openstack-swift | 09:47 | |
openstackgerrit | lokesh s proposed openstack/swift: py3: Fixes encoding and type error https://review.openstack.org/345520 | 09:47 |
kota_ | onovy: in the patch 351564, I addressed the lack of backword compatiblity with the newest PyECLib and liberasurecode < 1.2.0 | 09:49 |
patchbot | kota_: https://review.openstack.org/#/c/351564/ - pyeclib - Add tox/requirements settings to pass gate job (MERGED) | 09:49 |
kota_ | onovy: becuase liberasurecode-dev in ubuntu package is still older than 1.2.0 | 09:49 |
kota_ | but if >=1.2.0 is available in the package repo, we'd like to change the code to force liberasurecode >= 1.2.0 as the requirements | 09:50 |
kota_ | onovy: that's why I'm wondering if you coud work on that. | 09:51 |
mahatic_ | kota_: sorry had a long meeting and then lunch. checking the test now | 09:52 |
kota_ | mahatic_: np, I'm just starting to make a comment for your reply :-) | 09:53 |
kota_ | mahatic_: or if you're avaialble, I could take time to talk with you right now ;-) | 09:53 |
kota_ | instead of converstaion in the gerrit. | 09:53 |
mahatic_ | kota_: sure go ahead | 09:53 |
kota_ | ok | 09:53 |
*** furlongm has quit IRC | 09:54 | |
kota_ | so at the start point, I'm wondering why we need to set reclaim_age for get_hashes from auditor. | 09:54 |
kota_ | that's start point. | 09:54 |
*** furlongm has joined #openstack-swift | 09:54 | |
kota_ | because it should be in conf. | 09:54 |
kota_ | but | 09:54 |
kota_ | as you know, the reclaim_age is configured as for *replicator* | 09:55 |
*** asettle has quit IRC | 09:55 | |
kota_ | so you are trying to get the value from replicator, and then looking for the way how to pass it to diskfile | 09:55 |
kota_ | ritht? | 09:55 |
kota_ | right? | 09:55 |
*** asettle has joined #openstack-swift | 09:55 | |
mahatic_ | no | 09:56 |
mahatic_ | I want the conf value from auditor | 09:56 |
kota_ | ok | 09:56 |
mahatic_ | kota_: ^ | 09:56 |
kota_ | so you want to set different value from replicator's one as reclaim_age? | 09:57 |
mahatic_ | kota_: did you see my comment on auditor.py? I want the conf from auditor and if it's not specified fallback to replicator's | 09:57 |
mahatic_ | kota_: this is an alternate to moving it to DEFAULT | 09:57 |
kota_ | now looking | 09:58 |
mahatic_ | alternative* | 09:58 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 09:58 |
*** d0ugal has quit IRC | 09:58 | |
*** mvk has joined #openstack-swift | 09:59 | |
kota_ | that gives me another question though, why we need to set another value with replicator | 09:59 |
kota_ | is that safe? how large (or small) value is set, we're expecting? | 09:59 |
kota_ | ah, rsync_timout has same issue too. | 10:00 |
kota_ | ok, probably we have a couple of points to discuss. | 10:01 |
openstackgerrit | Christian Schwede proposed openstack/python-swiftclient: Make options arguments truely optional https://review.openstack.org/357063 | 10:01 |
kota_ | 1. we need to enable the reclaime_age get overwrite at auditor value | 10:01 |
openstackgerrit | Christian Schwede proposed openstack/swift: Deprecate swift-temp-url https://review.openstack.org/348173 | 10:01 |
kota_ | 2. how we can pass the reclaim_age value to DiskFileManager | 10:02 |
kota_ | the first still needs discussion because rsync_timeout has same issue. | 10:03 |
kota_ | what I'd like to tell you is latter one. | 10:03 |
kota_ | the second, | 10:03 |
*** d0ugal has joined #openstack-swift | 10:05 | |
kota_ | DiskFileManager already has reclaim_age for the instance variable, https://github.com/openstack/swift/blob/master/swift/obj/diskfile.py#L542 | 10:05 |
mahatic_ | kota_: I'm not sure why it wouldn't be safe? setting a value for auditor as well as replicator | 10:05 |
*** mvk has quit IRC | 10:05 | |
mahatic_ | kota_: for second one, correct it has a value, but I want to pass auditor's if it is specified | 10:06 |
mahatic_ | I think the latter one depends on an answer to the first | 10:06 |
kota_ | mahatic_: did you see the diff from me https://gist.github.com/bloodeagle40234/3d05f8cd4c217fc62581163e3a8e3078 | 10:07 |
kota_ | ? | 10:07 |
mahatic_ | kota_: yes, doesn't it directly make use of replicator's (and otherwise a default value)? | 10:08 |
kota_ | if you want to set the value from auditor conf, the diff still works because we can set the value at the init of DiskFileManager (via init for DiskFileRouter) | 10:08 |
mahatic_ | hmm, but why do we want to update the conf? | 10:09 |
kota_ | good question | 10:09 |
kota_ | the conf is a dict of key value pairs for "object-auditor" section | 10:09 |
mahatic_ | right | 10:10 |
openstackgerrit | Davanum Srinivas (dims) proposed openstack/swift: [WIP] Testing latest u-c https://review.openstack.org/318441 | 10:10 |
kota_ | so we don't have the reclaim_age in the conf dict | 10:10 |
mahatic_ | kota_: what if an op wants to set a different value? | 10:11 |
*** takashi has quit IRC | 10:11 | |
mahatic_ | kota_: how is the diff making a difference to the existing code? | 10:11 |
kota_ | mahatic_: set the value in auditor section because you already coded as auditor section prior to replicator https://gist.github.com/bloodeagle40234/3d05f8cd4c217fc62581163e3a8e3078#file-quick_fix_without_get_hashes_change-diff-L15 | 10:12 |
kota_ | the biggest difference is we don't need to add reclaim_age arg to get_hashes anymore (I don't do that in the diff though) | 10:13 |
mahatic_ | kota_: ok, so you want to avoid passing reclaim_age to get_hashes | 10:13 |
mahatic_ | kota_: right :) | 10:14 |
kota_ | https://gist.github.com/bloodeagle40234/3d05f8cd4c217fc62581163e3a8e3078#file-quick_fix_without_get_hashes_change-diff-L27 | 10:14 |
kota_ | just I'd like to make the change small enough | 10:14 |
mahatic_ | kota_: got it, thanks | 10:15 |
kota_ | mahatic_: yeah, I'd also say thank you for you. You're working a lot to address my comments. | 10:16 |
acoles | kota_: hi, re patch 355683, I just left a comment...if you have chance to change it today then I can vote before I leave for vacation | 10:18 |
patchbot | acoles: https://review.openstack.org/#/c/355683/ - swift - Follow up delayed reap probe test | 10:18 |
mahatic_ | kota_: np at all. I wanted to let you know that, I didn't add one of the tests you proposed. this one: | 10:18 |
mahatic_ | https://gist.github.com/bloodeagle40234/3224f3ccb7d78862cb35dec856c263bb#file-invalidate_hash_and_get_hashes_test-diff-L65 | 10:18 |
kota_ | acoles: thanks a lot! I'm going to do that soon! | 10:18 |
mahatic_ | it fails and needs some changes, will look at that. IIRC it doesn't do what it says (Sorry for being vague, will take a look again) | 10:19 |
*** mvk has joined #openstack-swift | 10:20 | |
acoles | kota_: ok. I think a separate patch to discuss using resetswift on success will be good. | 10:20 |
kota_ | mahatic: hmm, interesting. but it's ok to drop my work anyway to forcus on only the fix for current problem. | 10:22 |
kota_ | interesting on why that test failed :P | 10:22 |
mahatic_ | heh | 10:23 |
openstackgerrit | Kota Tsuyuzaki proposed openstack/swift: Follow up delayed reap probe test https://review.openstack.org/355683 | 10:26 |
kota_ | acoles: done^^ | 10:26 |
acoles | kota_: +2 | 10:27 |
kota_ | acoles: thanks! | 10:28 |
kota_ | acoles: ah, btw, I could work on the rebase on the optimistic GET again while you're on vacation? | 10:29 |
kota_ | can I? | 10:29 |
acoles | kota_: yes, please do. I'm sorry I have not yet had chance to review it again myself. | 10:33 |
acoles | kota_: I may get chance to do simple rebase today, if not then you can, thanks | 10:33 |
kota_ | acoles: ok, will do if you coudln't. I also didn't too sorry but I think I can do that in the next week. | 10:34 |
kota_ | didn't review yet sorry, i mean | 10:34 |
openstackgerrit | Kazuhiro MIYAHARA proposed openstack/swift: WIP: Swift Automated Tiering https://review.openstack.org/287057 | 10:45 |
*** m_kazuhiro has quit IRC | 10:52 | |
openstackgerrit | Kota Tsuyuzaki proposed openstack/swift: WIP: Cleanup probe test environment after running successfully https://review.openstack.org/357101 | 10:53 |
openstackgerrit | Merged openstack/swift: monkey_patch_mimetools() now does nothing on py3 https://review.openstack.org/347498 | 10:53 |
kota_ | the code for probe environment cleanup was divided and made as general | 10:53 |
kota_ | (not only for account reaper) | 10:54 |
openstackgerrit | Alistair Coles proposed openstack/swift: py3: Fixes encoding and type error https://review.openstack.org/345520 | 11:17 |
*** asettle has quit IRC | 11:25 | |
*** asettle has joined #openstack-swift | 11:26 | |
*** kei_yama_ has quit IRC | 11:28 | |
*** nikivi has joined #openstack-swift | 11:29 | |
*** SkyRocknRoll has quit IRC | 11:55 | |
*** tqtran has joined #openstack-swift | 11:56 | |
*** SkyRocknRoll has joined #openstack-swift | 11:57 | |
*** SkyRocknRoll has quit IRC | 11:58 | |
*** tqtran has quit IRC | 12:00 | |
*** mvk has quit IRC | 12:21 | |
*** mingyu has quit IRC | 12:32 | |
*** asettle has quit IRC | 12:34 | |
*** asettle has joined #openstack-swift | 12:34 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 12:41 |
onovy | kota_: it's already done | 12:53 |
onovy | https://packages.debian.org/sid/liberasurecode-dev | 12:53 |
onovy | https://packages.debian.org/jessie-backports/liberasurecode-dev | 12:53 |
*** mingyu has joined #openstack-swift | 12:54 | |
onovy | debian != ubuntu | 12:54 |
onovy | and pyeclib is updated too: https://packages.debian.org/jessie-backports/python-pyeclib | 12:54 |
*** klamath has joined #openstack-swift | 13:08 | |
*** klamath has quit IRC | 13:08 | |
*** klamath has joined #openstack-swift | 13:09 | |
openstackgerrit | Alistair Coles proposed openstack/swift: Follow up delayed reap probe test https://review.openstack.org/355683 | 13:10 |
acoles | kota_: I fixed pep8 error ^^ | 13:10 |
*** airdisa has joined #openstack-swift | 13:11 | |
*** links has quit IRC | 13:16 | |
*** thumpba has joined #openstack-swift | 13:17 | |
*** adu has joined #openstack-swift | 13:19 | |
*** ametts has joined #openstack-swift | 13:20 | |
*** joeljwright has quit IRC | 13:29 | |
*** dmorita has joined #openstack-swift | 13:30 | |
*** SkyRocknRoll has joined #openstack-swift | 13:31 | |
*** dmorita has quit IRC | 13:35 | |
*** BigWillie has joined #openstack-swift | 13:39 | |
*** joeljwright has joined #openstack-swift | 13:44 | |
*** ChanServ sets mode: +v joeljwright | 13:44 | |
*** adu has quit IRC | 13:53 | |
*** baojg has joined #openstack-swift | 13:55 | |
*** baojg has quit IRC | 13:56 | |
*** ppai has joined #openstack-swift | 14:01 | |
*** asettle has quit IRC | 14:20 | |
*** asettle has joined #openstack-swift | 14:21 | |
*** asettle has quit IRC | 14:21 | |
*** asettle has joined #openstack-swift | 14:21 | |
openstackgerrit | Janie Richling proposed openstack/swift: Add test for POST to DLO manifest file https://review.openstack.org/347545 | 14:32 |
jrichli | acoles ^^ should be fixed | 14:32 |
acoles | jrichli: aww, you put the oo back into foo | 14:34 |
jrichli | :-p | 14:34 |
jrichli | well, i had to split on the next version of it anyway with new_x_o_m | 14:34 |
*** d0ugal has quit IRC | 14:41 | |
*** ppai has quit IRC | 14:46 | |
*** d0ugal has joined #openstack-swift | 14:48 | |
*** d0ugal has quit IRC | 14:48 | |
*** d0ugal has joined #openstack-swift | 14:48 | |
*** dmellado|off is now known as dmellado | 14:52 | |
*** bill_az has joined #openstack-swift | 14:54 | |
jrichli | does anyone have any ideas how we could fix Bug 1566124 - or should we just document the discrepancy? | 14:56 |
openstack | bug 1566124 in OpenStack Object Storage (swift) "Version overwrite of an object with max length name fails" [Undecided,New] https://launchpad.net/bugs/1566124 | 14:56 |
*** bill_az has quit IRC | 15:01 | |
*** d0ugal has quit IRC | 15:09 | |
*** klrmn has joined #openstack-swift | 15:12 | |
*** mvk has joined #openstack-swift | 15:19 | |
*** rcernin has quit IRC | 15:22 | |
*** d0ugal has joined #openstack-swift | 15:24 | |
notmyname | good morning | 15:32 |
*** vint_bra has joined #openstack-swift | 15:37 | |
*** joeljwright has quit IRC | 15:45 | |
*** mvk has quit IRC | 15:50 | |
*** Suyi has joined #openstack-swift | 15:54 | |
*** tqtran has joined #openstack-swift | 15:57 | |
*** tqtran has quit IRC | 16:02 | |
*** joeljwright has joined #openstack-swift | 16:12 | |
*** ChanServ sets mode: +v joeljwright | 16:12 | |
*** arch-nemesis has joined #openstack-swift | 16:12 | |
*** ametts has quit IRC | 16:16 | |
*** ametts has joined #openstack-swift | 16:21 | |
*** zaitcev has joined #openstack-swift | 16:22 | |
*** ChanServ sets mode: +v zaitcev | 16:22 | |
*** rledisez has quit IRC | 16:25 | |
*** tesseract- has quit IRC | 16:26 | |
*** dronshaw has joined #openstack-swift | 16:30 | |
*** dronshaw has left #openstack-swift | 16:32 | |
*** dmorita has joined #openstack-swift | 16:35 | |
*** thumpba_ has joined #openstack-swift | 16:37 | |
*** nikivi has quit IRC | 16:38 | |
*** thumpba has quit IRC | 16:39 | |
*** SkyRocknRoll has quit IRC | 16:40 | |
*** jordanP has quit IRC | 16:43 | |
*** nikivi has joined #openstack-swift | 16:56 | |
*** asettle has quit IRC | 16:58 | |
*** asettle has joined #openstack-swift | 16:58 | |
*** joeljwright has quit IRC | 16:59 | |
*** jordanP has joined #openstack-swift | 17:02 | |
*** asettle has quit IRC | 17:03 | |
timburke | good morning | 17:08 |
*** ober37` has joined #openstack-swift | 17:12 | |
*** ober37 has quit IRC | 17:13 | |
*** mminesh has quit IRC | 17:13 | |
*** bkeller` has quit IRC | 17:13 | |
*** ober37` has quit IRC | 17:13 | |
*** ningliu has quit IRC | 17:13 | |
*** bkeller` has joined #openstack-swift | 17:13 | |
*** mminesh has joined #openstack-swift | 17:14 | |
*** ningliu has joined #openstack-swift | 17:15 | |
*** itlinux has joined #openstack-swift | 17:15 | |
*** jrichli` has joined #openstack-swift | 17:16 | |
*** ober37 has joined #openstack-swift | 17:16 | |
*** cdelatte has joined #openstack-swift | 17:19 | |
*** furlongm has quit IRC | 17:24 | |
*** furlongm has joined #openstack-swift | 17:24 | |
*** klrmn has quit IRC | 17:27 | |
*** klrmn has joined #openstack-swift | 17:31 | |
*** tqtran has joined #openstack-swift | 17:39 | |
*** rcernin has joined #openstack-swift | 17:51 | |
openstackgerrit | Alistair Coles proposed openstack/swift: Add process level concurrency to container sync https://review.openstack.org/210099 | 17:52 |
* acoles will be back in September | 17:58 | |
*** acoles is now known as acoles_ | 17:59 | |
*** mingyu has quit IRC | 18:13 | |
*** furlongm has quit IRC | 18:15 | |
*** furlongm has joined #openstack-swift | 18:15 | |
*** manous has joined #openstack-swift | 18:16 | |
*** zul has quit IRC | 18:16 | |
*** manous has quit IRC | 18:21 | |
*** manous has joined #openstack-swift | 18:22 | |
*** zul has joined #openstack-swift | 18:23 | |
*** manous has quit IRC | 18:27 | |
*** jrichli has quit IRC | 18:30 | |
*** jrichli` is now known as jrichli | 18:30 | |
*** manous has joined #openstack-swift | 18:31 | |
jrichli | bkeller` set me up with an account on his bouncer, so I have still successfully avoided having to own one :-p | 18:48 |
*** catintheroof has joined #openstack-swift | 18:49 | |
*** nikivi has quit IRC | 18:59 | |
*** itlinux has quit IRC | 19:02 | |
*** fifieldt has quit IRC | 19:06 | |
clayg | irssi in a screen on a cloud vm ftw - bouncers are far too complicated and desktop irc clinets are over-rated | 19:10 |
tqtran | hello, does anyone know why we check for the acl attribute in request objects? http://paste.openstack.org/show/560848/ did a search for it, seems like we are expecting users to pass this in? | 19:12 |
tqtran | or is there some magic somewhere that i am not seeing? | 19:12 |
timburke | tqtran: the acl gets set in the proxy controllers, for example https://github.com/openstack/swift/blob/2.9.0/swift/proxy/controllers/obj.py#L174 | 19:17 |
*** fifieldt has joined #openstack-swift | 19:17 | |
timburke | in all the controllers, it should happen shortly before swift.authorize callback is called | 19:18 |
*** thumpba_ has quit IRC | 19:29 | |
*** thumpba has joined #openstack-swift | 19:39 | |
*** bill_az has joined #openstack-swift | 19:48 | |
*** thumpba has quit IRC | 20:04 | |
*** gyee has joined #openstack-swift | 20:11 | |
jrichli | timburke: thanks for your feedback on Bug 1566124. Ill start on something along those lines | 20:13 |
openstack | bug 1566124 in OpenStack Object Storage (swift) "Version overwrite of an object with max length name fails" [Undecided,New] https://launchpad.net/bugs/1566124 | 20:13 |
timburke | jrichli: thanks! and i'm taking a look at https://review.openstack.org/#/c/356239 now | 20:14 |
patchbot | timburke: patch 356239 - swift - Allow DLO manifest to be updated when using post-a... | 20:14 |
jrichli | clayg: good to know. ill have to look into that | 20:15 |
jrichli | timburke: great, thanks! in case you didnt see, there is a simple parent review for that too | 20:16 |
*** psachin has joined #openstack-swift | 20:20 | |
*** BigWillie has quit IRC | 20:24 | |
*** diogogmt has joined #openstack-swift | 20:27 | |
*** Jeffrey4l has quit IRC | 20:35 | |
*** Jeffrey4l has joined #openstack-swift | 20:35 | |
*** nadeem has joined #openstack-swift | 20:41 | |
*** nadeem has quit IRC | 20:41 | |
*** nadeem has joined #openstack-swift | 20:42 | |
clayg | irssi in a screen on a cloud vm ftw - bouncers are far too complicated and desktop irc clinets are over-rated | 20:54 |
clayg | sorry | 20:54 |
*** cdelatte has quit IRC | 21:16 | |
*** catintheroof has quit IRC | 21:21 | |
csmart | morning | 21:32 |
tqtran | timburke: ah thanks! | 21:32 |
*** psachin has quit IRC | 21:44 | |
*** furlongm has quit IRC | 21:45 | |
*** furlongm has joined #openstack-swift | 21:46 | |
*** nadeem has quit IRC | 22:01 | |
mattoliverau | Morning | 22:08 |
*** vint_bra has quit IRC | 22:14 | |
*** klamath has quit IRC | 22:22 | |
*** asettle has joined #openstack-swift | 22:24 | |
*** asettle has quit IRC | 22:32 | |
*** ametts has quit IRC | 22:51 | |
openstackgerrit | Samuel Merritt proposed openstack/swift: Support multi-range GETs for static large objects. https://review.openstack.org/352657 | 22:59 |
openstackgerrit | Daisuke Morita proposed openstack/swift: WIP: Changing Policies https://review.openstack.org/209329 | 23:09 |
*** jordanP has quit IRC | 23:16 | |
*** chlong has quit IRC | 23:20 | |
openstackgerrit | Merged openstack/swift: Add test for POST to DLO manifest file https://review.openstack.org/347545 | 23:25 |
*** kei_yama has joined #openstack-swift | 23:26 | |
*** rcernin has quit IRC | 23:32 | |
*** airdisa has quit IRC | 23:35 | |
*** hosanai has joined #openstack-swift | 23:43 | |
*** ChanServ sets mode: +v hosanai | 23:43 | |
*** ouchkernel has quit IRC | 23:48 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 23:51 |
*** kei_yama has quit IRC | 23:53 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!