15:57:01 #startmeeting 15:57:01 Meeting started Fri Jan 10 15:57:01 2014 UTC. The chair is johnmark. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:57:01 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:57:01 Meeting started Fri Jan 10 16:06:13 2014 UTC. The chair is johnmark. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:57:02 Useful Commands: #action #agreed #help #info #idea #link #topic. 15:57:11 #topic hangout with semiosis 15:58:33 yes? 15:59:08 * kkeithley is not on the hangout! 16:01:51 wha? 16:01:58 kkeithley: you got the invite, right? 16:02:01 lol 16:02:42 actually, no, I did not. Where did you send it? 16:03:51 hrm... maybe there's another kaleb keithley? 16:06:29 oh, you sent it to my kkeithleatredhat@gmail address 16:06:31 doh 16:08:59 #link https://forge.gluster.org/libgfapi-jni 16:09:15 #link https://forge.gluster.org/glusterfs-java-filesystem 16:14:14 kkeithley: doh :( 16:18:49 john mark, have you been talking to any companies with well known java applications about using this behind their applications? I think it would help drive market adoption in the enterprise. 16:18:50 * kkeithley just linked my two gmail accounts 16:19:22 dbruhn: great question. once we get it into a more releasable state, that' sa good idea 16:19:35 I know a couple of the hadoop integration guys are interested 16:20:15 Would redhat/gluster be willing to assign coding resources to help those integrations along? 16:20:55 semiosis, how is the performance compared to NFS/Fuse? 16:29:45 johnmark: Did you ask abt the performance compared to NFS/Fuse? 16:31:08 jclift: he did 16:31:26 jclift: ja 16:31:46 jclift, he said there hasn't been any performance testing, due to functional development taking precedence. 16:32:10 but he assumed that it would be faster than fuse, if fuse was a bottleneck for you 16:34:10 we would expect it to be faster, e.g., like Samba using the gfapi vfs plug-in, wouldn't we? 16:34:46 or libvirt with gfapi. 16:34:48 kkeithley: I would assume 16:35:08 but you know what they say about "assume" 16:36:09 indeed 16:36:12 depends on the nature of workload that you are using .. note that libgfapi is entirely in userspace and hence you miss out on caching that happens in the kernel 16:37:17 however samba with vfs_plugin has shown significant improvement over samba with fuse - again I "assume" that coming through libgfapi would be beneficial. 16:37:38 right 16:41:19 johnmark: How many live viewers? 16:41:26 jclift: the max was 16 16:41:36 and there were >10 the whole time 16:41:41 #endmeeting