Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2017-02-14

| Channels | #gluster-dev index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:20 cholcombe nixpanic_, are the glfs-handles., glfs-internal.h part of the public gfapi?
00:20 cholcombe it looks like i need to update my rust crate.  I don't have those in there
00:32 ankit joined #gluster-dev
00:36 gyadav_ joined #gluster-dev
00:55 gem joined #gluster-dev
01:38 ankit joined #gluster-dev
02:17 riyas joined #gluster-dev
02:26 gem joined #gluster-dev
02:28 Shu6h3ndu joined #gluster-dev
02:34 skoduri joined #gluster-dev
02:48 ilbot3 joined #gluster-dev
02:48 Topic for #gluster-dev is now Gluster Development Channel - http://gluster.org | For general chat go to #gluster | Patches - http://review.gluster.org/ | Channel Logs - https://botbot.me/freenode/gluster-dev/ & http://irclog.perlgeek.de/gluster-dev/
03:39 nbalacha joined #gluster-dev
03:46 vimal joined #gluster-dev
03:58 magrawal joined #gluster-dev
03:59 aravindavk joined #gluster-dev
04:03 gyadav_ joined #gluster-dev
04:18 nbalacha joined #gluster-dev
04:19 skumar joined #gluster-dev
04:25 vimal joined #gluster-dev
04:41 skumar joined #gluster-dev
04:48 skoduri joined #gluster-dev
04:49 riyas joined #gluster-dev
04:59 rafi joined #gluster-dev
05:06 ppai joined #gluster-dev
05:15 ndarshan joined #gluster-dev
05:16 lpabon joined #gluster-dev
05:17 mchangir joined #gluster-dev
05:20 Humble joined #gluster-dev
05:21 gem joined #gluster-dev
05:24 skumar_ joined #gluster-dev
05:25 karthik_us joined #gluster-dev
05:33 kdhananjay joined #gluster-dev
05:39 susant joined #gluster-dev
05:45 jiffin joined #gluster-dev
05:47 mchangir joined #gluster-dev
05:52 riyas joined #gluster-dev
05:53 k4n0 joined #gluster-dev
05:57 Guest37230 joined #gluster-dev
05:58 prasanth joined #gluster-dev
06:06 ashiq joined #gluster-dev
06:08 sanoj joined #gluster-dev
06:08 susant joined #gluster-dev
06:08 jiffin1 joined #gluster-dev
06:20 ankit_ joined #gluster-dev
06:24 sanoj joined #gluster-dev
06:33 hgowtham joined #gluster-dev
06:35 Humble joined #gluster-dev
06:40 mchangir joined #gluster-dev
06:53 hgowtham joined #gluster-dev
07:03 kotreshhr joined #gluster-dev
07:07 msvbhat joined #gluster-dev
07:17 jiffin1 joined #gluster-dev
07:22 xavih ndevos, anoopcs: sorry, I wasn't here yesterday...
07:23 xavih ndevos, anoopcs: I think that selinux requires that if a file is mmaped with execution rights, it must be created in a directory with bin_t type
07:23 xavih ndevos, anoopcs: otherwise it refuses to mmap
07:23 anoopcs xavih, Yes.,
07:23 xavih ndevos, anoopcs: will the temporary directory have this type ?
07:24 xavih ndevos, anoopcs: I already tried /var/run and some others, but none of them worked by default
07:25 ndevos xavih: if /var/run/gluster does not work, we need to create /var/run/gluster/lib and set the correct contexts there, that wont be a big issue
07:26 xavih ndevos: if that's easy to, it's ok for me :)
07:27 xavih ndevos: I don't know selinux, so I wanted to use a directory that already had this type
07:27 xavih ndevos: how is this done ?
07:27 ndevos xavih: it should not be too hard, and it would be a much better location :)
07:28 ndevos xavih: I think anoopcs is looking into that (and if he isnt, I need to know)
07:28 xavih ndevos: ah, ok :)
07:29 Humble joined #gluster-dev
07:30 xavih anoopcs: if you need any help, please ask :)
07:32 anoopcs xavih, I hope you have gone through https://bugzilla.redhat.com​/show_bug.cgi?id=1421649#c4
07:32 glusterbot Bug 1421649: unspecified, unspecified, ---, bugs, NEW , When using a fuse mount for client, EC volumes do not mount.
07:33 anoopcs xavih, What is your opinion?
07:34 ndevos xavih, anoopcs: as mentioned in the bug, I want EC also to work when a non-root user runs a gfapi application... that one is more tricky
07:34 anoopcs Yeah...
07:34 xavih anoopcs: yes, I have seen it
07:35 xavih ndevos, anoopcs: I can do that, but it's fine that an application itself detects the user it's running, if selinux is enabled, automatically create a directory and set selinux types ?
07:36 xavih ndevos, anoopcs: it seems like a lot of troubles only for a specific OS
07:37 xavih ndevos, anoopcs: isn't there any other way to do that ? I mean, to have something similar to /tmp where many users can create files but with the right selinux type created by the installation scripts...
07:38 anoopcs xavih, ndevos, In order to avoid complexities I think it better to use /tmp and get around with selinux using proper selinux policy
07:38 ndevos xavih: I do not think we should create a world-writable directory with the bin_t context, a bit like /tmp, I guess the security folks would be against it
07:39 ndevos hmm, isnt siddarth (our security guy) in this channel?
07:40 xavih ndevos: so the best option would be to create a new directory dynamically under /var/user/... and set the bin_t type programmatically ?
07:40 xavih ndevos: if that's the best option, I can do that
07:40 anoopcs ndevos, I don't think he is...
07:40 xavih ndevos: sorry, /run/user/...
07:41 ndevos xavih: close, but I think the bin_t type should be set from somewhere else, maybe the selinux policy can set types when a directory is created (I dont know though)
07:42 BatS9 joined #gluster-dev
07:43 xavih ndevos: you are right. I think I saw something about this...
07:43 * ndevos poked siddarth, he might join in a bit
07:44 xavih ndevos: /etc/selinux/targeted/contexts/files/file_contexts
07:45 ndevos xavih: yes, like that
07:45 xavih ndevos: maybe we could add a rule to set only some files, not all files inside a directory
07:46 ndevos xavih: probably, but lets get some details from some people that know selinux better than us :)
07:46 xavih ndevos: right :)
07:47 ndevos anoopcs: could you put a needinfo on siddarth in the bug? he's not on freenode atm
07:56 Humble joined #gluster-dev
08:06 kdhananjay joined #gluster-dev
08:06 nishanth joined #gluster-dev
08:16 ndevos xavih, anoopcs: is this a problem only against master, or also for 3.10?
08:16 xavih ndevos: it's also merged in 3.10
08:17 ndevos xavih: oh :-(
08:17 xavih ndevos: yes :(
08:18 ndevos xavih: should we suggest to block the 3.10 release for this?
08:19 xavih ndevos: not sure. There's a workaround by disabling the use of dynamic code generation
08:19 xavih ndevos: dynamic code can fail if the directory doesn't exist or it's not run as root
08:20 ndevos xavih: oh, ok, not a blocker then, but it requires a mount/volume-option?
08:21 xavih ndevos: it requires to set the volume option "disperse.cpu-extensions" to "none"
08:21 xavih ndevos: let me double check though...
08:21 ndevos xavih: ok, we need to get that in the release notes under "known issues" then
08:23 ndevos xavih: would it be reasonable to change the behaviour to fallback to the non-dynamic code in case of a mmap() error?
08:24 atinm joined #gluster-dev
08:24 xavih ndevos: that would be a good thing, not only for mmap() but also for any other possible error generated during the dynamic code processing...
08:24 xavih ndevos: I'll do that
08:25 ndevos xavih++ awesome, that sounds like a good solution until we know where/how to place the tmp files under SELinux
08:25 glusterbot ndevos: xavih's karma is now 26
08:26 xavih ndevos: do I use the same bug for this change or do I file a new bug report ? what's better ?
08:29 ndevos xavih: create a new bug for mainline and clone it to 3.10, we still need to address the other difficulties too, but they can be done seperate
08:29 xavih ndevos: ok, thanks :)
08:31 ndevos xavih: in the 3.10 clone, add "glusterfs-3.10.0" to the "blocks" field so that shyam and kkeithley can track it easily
08:36 nbalacha joined #gluster-dev
08:41 suliba joined #gluster-dev
08:47 jiffin1 joined #gluster-dev
08:49 skoduri joined #gluster-dev
09:07 anoopcs xavih, ndevos : Seems like you guys have figured out an interim solution.. Thanks
09:09 anoopcs And as mentioned above by ndevos, a fallback to non-dynamic code approach looks good to me.
09:13 anoopcs ndevos++ for helping out with your suggestions...
09:13 glusterbot anoopcs: ndevos's karma is now 331
09:13 anoopcs xavih++ and you too.
09:13 glusterbot anoopcs: xavih's karma is now 27
09:15 ShwethaHP joined #gluster-dev
09:30 jiffin1 joined #gluster-dev
09:31 poornima joined #gluster-dev
09:32 Saravanakmr joined #gluster-dev
09:35 ndevos anoopcs: interim solution yes, but we still need input from siddarth on the bz for a proper solution
09:36 anoopcs ndevos, Yes...But have you already raised a needinfo?
09:39 xavih ndevos, anoopcs: I've created the fallback patch. I'm doing some tests before uploading it
09:39 anoopcs xavih, Cool.
09:39 anoopcs \o/
09:40 gem joined #gluster-dev
09:40 xavih ndevos++: thanks for the idea :)
09:40 glusterbot xavih: ndevos's karma is now 332
10:05 percevalbot joined #gluster-dev
10:41 mchangir joined #gluster-dev
11:01 mchangir joined #gluster-dev
11:27 skoduri joined #gluster-dev
11:38 mchangir joined #gluster-dev
11:42 kotreshhr left #gluster-dev
11:49 susant left #gluster-dev
11:55 hgowtham REMINDER: gluster community bug triage in ~5 minutes at #gluster-dev
11:59 ndevos shyam, kkeithley: https://bugzilla.redhat.com​/show_bug.cgi?id=1421956#c1 would be a blocker for 3.10.0, a patch is under review for the master branch
11:59 glusterbot Bug 1421956: high, unspecified, ---, xhernandez, ASSIGNED , Disperse: Fallback to pre-compiled code execution when dynamic code generation fails
12:22 bfoster joined #gluster-dev
12:26 nishanth joined #gluster-dev
12:30 rraja joined #gluster-dev
12:38 hgowtham ndevos++
12:38 glusterbot hgowtham: ndevos's karma is now 333
12:53 ndevos hgowtham++ too :)
12:53 glusterbot ndevos: hgowtham's karma is now 55
12:57 rraja joined #gluster-dev
13:22 skoduri joined #gluster-dev
13:26 atm0sphere joined #gluster-dev
13:57 rraja joined #gluster-dev
14:00 ankitr joined #gluster-dev
14:05 k4n0 joined #gluster-dev
14:08 jiffin joined #gluster-dev
14:14 mchangir joined #gluster-dev
14:16 riyas joined #gluster-dev
14:19 ppai joined #gluster-dev
14:19 atinm joined #gluster-dev
14:34 gem joined #gluster-dev
14:40 gem joined #gluster-dev
14:46 PotatoGim joined #gluster-dev
14:54 skumar joined #gluster-dev
14:54 nbalacha joined #gluster-dev
14:58 ppai joined #gluster-dev
15:22 skumar joined #gluster-dev
15:24 skumar joined #gluster-dev
15:30 rraja joined #gluster-dev
15:32 shyam joined #gluster-dev
15:32 shyam ndevos: ack, saw the tracker bug updated as well!
16:05 wushudoin joined #gluster-dev
16:05 msvbhat joined #gluster-dev
16:18 lpabon joined #gluster-dev
16:51 jiffin joined #gluster-dev
16:55 ShwethaHP joined #gluster-dev
16:55 aravindavk joined #gluster-dev
17:15 skumar_ joined #gluster-dev
17:31 ndevos shyam: I just notice that there is no op-version for 3.10. Are there no volume options added in the release that require those?
17:39 ndevos shyam: oh, scratch that, I thought I did a rebase of a patch, but obviously not :-/
17:44 skumar joined #gluster-dev
19:58 vbellur joined #gluster-dev
20:01 cholcombe joined #gluster-dev
20:05 cholcombe joined #gluster-dev
20:10 msvbhat joined #gluster-dev
20:14 cholcombe joined #gluster-dev
20:52 vbellur joined #gluster-dev
22:38 overclk_ joined #gluster-dev
22:43 nthomas joined #gluster-dev
22:46 cholcombe joined #gluster-dev
22:49 [o__o] joined #gluster-dev
22:50 PotatoGim joined #gluster-dev
22:51 wushudoin joined #gluster-dev
22:52 shaunm joined #gluster-dev
22:52 jtc joined #gluster-dev
23:12 shyam joined #gluster-dev
23:24 foster joined #gluster-dev
23:36 vbellur joined #gluster-dev
23:58 anoopcs joined #gluster-dev

| Channels | #gluster-dev index | Today | | Search | Google Search | Plain-Text | summary