Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2015-07-21

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

All times shown according to UTC.

Time Nick Message
01:54 dlambrig left #gluster-dev
03:12 dlambrig joined #gluster-dev
03:21 jarrpa joined #gluster-dev
03:25 schandra joined #gluster-dev
03:52 itisravi joined #gluster-dev
03:52 spandit joined #gluster-dev
03:55 shubhendu joined #gluster-dev
04:01 atinm joined #gluster-dev
04:12 soumya joined #gluster-dev
04:15 dlambrig joined #gluster-dev
04:16 josferna joined #gluster-dev
04:35 nbalacha joined #gluster-dev
04:39 dlambrig joined #gluster-dev
04:44 vimal joined #gluster-dev
04:46 ashishpandey joined #gluster-dev
04:48 sakshi joined #gluster-dev
05:00 gem joined #gluster-dev
05:04 rafi joined #gluster-dev
05:07 jiffin joined #gluster-dev
05:10 ndarshan joined #gluster-dev
05:12 pppp joined #gluster-dev
05:13 hagarth joined #gluster-dev
05:16 deepakcs joined #gluster-dev
05:21 soumya_ joined #gluster-dev
05:37 rjoseph joined #gluster-dev
05:41 Manikandan joined #gluster-dev
05:42 hgowtham joined #gluster-dev
05:45 ashiq joined #gluster-dev
05:49 Bhaskarakiran joined #gluster-dev
05:50 kanagaraj joined #gluster-dev
05:54 kdhananjay joined #gluster-dev
05:57 aravindavk joined #gluster-dev
06:01 asengupt joined #gluster-dev
06:03 vmallika joined #gluster-dev
06:04 deepakcs joined #gluster-dev
06:05 overclk joined #gluster-dev
06:07 anekkunt joined #gluster-dev
06:10 Manikandan joined #gluster-dev
06:17 jiffin joined #gluster-dev
06:18 shubhendu joined #gluster-dev
06:19 ndarshan joined #gluster-dev
06:21 byreddy joined #gluster-dev
06:22 jiffin left #gluster-dev
06:31 jiffin joined #gluster-dev
06:33 krishnan_p joined #gluster-dev
06:41 kshlm joined #gluster-dev
06:54 raghu joined #gluster-dev
06:54 Manikandan joined #gluster-dev
07:00 raghu overclk: can you retrigger regression for http://review.gluster.org/#/c/11703/? I am not able to login to jenkins
07:16 soumya_ joined #gluster-dev
07:23 Manikandan joined #gluster-dev
07:24 topshare joined #gluster-dev
07:35 atalur joined #gluster-dev
07:37 overclk raghu, sure.
07:42 hgowtham kshlm++
07:42 glusterbot hgowtham: kshlm's karma is now 20
07:44 kshlm raghu, you're still not able to log-in?
07:44 gbit joined #gluster-dev
07:53 shubhendu joined #gluster-dev
07:54 ndarshan joined #gluster-dev
07:56 soumya_ joined #gluster-dev
08:03 Manikandan joined #gluster-dev
08:05 topshare joined #gluster-dev
08:09 topshare joined #gluster-dev
08:10 topshare joined #gluster-dev
08:32 Bhaskarakiran joined #gluster-dev
08:35 topshare joined #gluster-dev
08:39 atalur xavih, ping. Shall we proceed?
08:39 xavih atalur: yes
08:40 xavih atalur: what do you think ?
08:41 atalur xavih, from the details about the framework that you had sent, I have a question.
08:41 atalur xavih, in the last part where you explain the framework from afr's perspective.
08:42 topshare joined #gluster-dev
08:42 atalur xavih, I'm unclear if you say txn-0 waits to send the request or not. till txn-1 receives and processes it
08:44 pranithk joined #gluster-dev
08:45 xavih atalur: txn-0 can start a non-blocking lock call only if no other txn-x has already failed a non-blocking call (this means that sequential blocking lock calls will be sent, so no need to send a non-blocking one). Otherwise it must wait until 2 requests have been seen by all txn-x xlators (as specified in gf_txn_prepare())
08:46 xavih atalur: when txn-1 receives the second request, it can also start a non-blocking lock call or it can initiate the sequential locks
08:47 xavih atalur: anyway, txn-xlators will also use library calls that will be aware of the global state of the transaction. It will be that library function who will issue the requests on behalf of the txn-xlators
08:48 xavih atalur: by "requests" I mean locking calls
08:51 gem joined #gluster-dev
08:52 atalur xavih, hmm. okay.
08:52 xavih atalur: I need to go for 15 minutes...
08:52 atalur xavih, okay. ping me once you come back. I'll be here
08:52 xavih atalur: ok
08:54 shubhendu joined #gluster-dev
09:03 xavih atalur: I'm here
09:05 kshlm joined #gluster-dev
09:06 jiffin joined #gluster-dev
09:10 kdhananjay joined #gluster-dev
09:11 atalur xavih, alright. one more question. from my understanding multiple inodes can be added to the same transaction. Is that correct?
09:12 xavih atalur: yes
09:12 xavih atalur: there are few fops that require that, though
09:12 gem joined #gluster-dev
09:12 atalur xavih, so that is just fop dependent right?
09:13 atalur xavih, I mean.  To succeed on a given brick, all inodes bound to that brick must be acquired successfully. If one fails, all others are released and the brick won't be used for the transaction.
09:13 atalur xavih, ^^ failing of one inode on a brick should lead to failure of all others in the brick only if that is a must, right?
09:14 xavih atalur: yes. If one fails, we cannot assure that the fop will be executed atomically with respect of other bricks. So we don't allow execution of fop if at least one lock fails
09:15 xavih atalur: a possible expansion would be to allow the user xlator to add custom inodes to the transaction, even if not needed for one fop. This would allow the execution of complex transactions that require multiple fops
09:16 xavih atalur: however this adds complexity and probably is not required right now
09:16 atalur xavih, the framework details say inode is added to the transaction by txn-xlator. How does txn-xlator decide it? Is it based on the information passed by the initiator xlator?
09:17 soumya joined #gluster-dev
09:17 shubhendu joined #gluster-dev
09:17 xavih atalur: it depends on the fop. A write will add the inode of the file written. A rename will add the inodes of the source and target directories
09:18 xavih atalur: it's implicit information contained in the request itself
09:18 atalur xavih, correct. yes that's what I wanted to confirm
09:22 topshare joined #gluster-dev
09:22 krishnan_p joined #gluster-dev
09:22 atinm joined #gluster-dev
09:23 pranithk joined #gluster-dev
09:23 ndevos atinm, rafi: who's doing the bug triage today?
09:26 atalur xavih, Let me get back to you in 10 minutes.
09:26 xavih atalur: sure
09:30 atinm ndevos, I've a meeting till 12:00 UTC and it might extend, that's the risk with me
09:30 atinm rafi, how about you?
09:31 ndevos atinm: is that a meeting where you need to be actively participating until the end, or could you leave early? :)
09:31 atinm ndevos, I can't guarantee that :(
09:34 atalur xavih, looks good till now. :) Although I'm yet to think of possible races that might be involved.
09:35 xavih atalur: I also need to work on it deeper, but I think there are two very interesting features: 1. Very easy to use api for user xlators, and 2. user xlator doesn't need to implement callbacks for transactions. It can create a transaction and immediately send the request
09:36 xavih atalur: this contributes to smaller and easy to read code
09:36 soumya joined #gluster-dev
09:36 rafi atinm: ndevos : there was a discussion to move our scrum meet to 5:30,
09:36 xavih atalur: at the same time it allows transaction nesting, that I think could be a very interesting feature for more complex graphs
09:38 ndevos rafi: if you already have an other meeting at that time, I guess scrum needs to wait ;-)
09:39 rafi ndevos: :)
09:39 rafi itisravi : can you host the meeting ?
09:40 itisravi rafi: I can with some help :)
09:40 atalur joined #gluster-dev
09:41 rafi itisravi: I will try to join on time or little late, ;)
09:41 atalur xavih, hmm, agreed. :)
09:41 itisravi rafi: okay.
09:41 rafi ndevos: if I couldn't make it on time , ndevos , can you help him ?
09:43 kdhananjay1 joined #gluster-dev
09:49 ndevos rafi, itisravi: sure, I'll try to be on time too, we have our scrum call 15 minutes before the bz triage
09:50 itisravi ndevos: rafi: okay
09:58 atinm joined #gluster-dev
09:59 krishnan_p joined #gluster-dev
10:13 kshlm joined #gluster-dev
10:13 ndevos csim: do you have an objection or suggestion for hosting NFS-Ganesha mailinglists on gluster.org? http://www.gluster.org/pipermail/gluster-infra/2015-July/001312.html
10:18 akay1 joined #gluster-dev
10:20 kkeithley1 joined #gluster-dev
10:34 jiffin joined #gluster-dev
10:35 pranithk joined #gluster-dev
10:38 soumya_ joined #gluster-dev
10:54 csim ndevos: no objection for me
10:55 csim csim: want me to do it, or you have access and ca n do it ?
10:55 csim ( still have lunch and a ansible helpdesk session, but I can do it in the afternoon, provided my laptop do not lose power )
11:12 atinm joined #gluster-dev
11:14 ira joined #gluster-dev
11:22 krishnan_p joined #gluster-dev
11:28 dlambrig joined #gluster-dev
11:35 ndevos csim: I can not do that, so please feel free to :)
11:35 itisravi joined #gluster-dev
11:41 topshare joined #gluster-dev
11:44 deepakcs joined #gluster-dev
11:46 firemanxbr joined #gluster-dev
11:56 soumya_ joined #gluster-dev
11:57 jiffin joined #gluster-dev
11:57 itisravi REMINDER: Gluster Community Bug Triage meeting starting in a few minutes in #gluster-meeting
12:01 rafi1 joined #gluster-dev
12:02 krishnan_p joined #gluster-dev
12:03 atinm joined #gluster-dev
12:06 hagarth joined #gluster-dev
12:06 sakshi joined #gluster-dev
12:11 jrm16020 joined #gluster-dev
12:16 overclk joined #gluster-dev
12:31 pranithk left #gluster-dev
12:32 rafi1 itisravi++
12:32 glusterbot rafi1: itisravi's karma is now 9
12:32 ndevos itisravi++
12:32 glusterbot ndevos: itisravi's karma is now 10
12:33 itisravi ndevos: thanks!
12:35 ndevos kkeithley_: please respond to the ganesha mailinglist email on infra@ :)
12:42 rafi joined #gluster-dev
12:55 nbalacha joined #gluster-dev
13:00 shaunm__ joined #gluster-dev
13:14 pppp joined #gluster-dev
13:19 ndevos hchiramm: we used to upload presentation on a wiki page, where do we put them now?
13:19 soumya joined #gluster-dev
13:22 shaunm_ joined #gluster-dev
13:25 kkeithley_ Where is the etherpad for "this week in Gluster" ?
13:25 ndevos https://public.pad.fsfe.org/p/gluster-weekly-news
13:26 kkeithley_ thanks
13:43 ndevos hmm, I cant remember what I was going to look at next... if I promised someone something, please ping that again :)
13:43 soumya_ joined #gluster-dev
13:49 kkeithley_ not the SELinux issue?
13:52 nbalacha joined #gluster-dev
13:59 ndevos no, that wasnt it...
14:04 wushudoin joined #gluster-dev
14:16 topshare joined #gluster-dev
14:19 wushudoin joined #gluster-dev
14:33 kanagaraj joined #gluster-dev
14:37 kanagaraj joined #gluster-dev
14:42 kanagaraj joined #gluster-dev
14:44 lpabon joined #gluster-dev
14:44 lkoranda joined #gluster-dev
14:45 pousley joined #gluster-dev
15:00 pousley_ joined #gluster-dev
15:09 sankarsh` joined #gluster-dev
15:15 vimal joined #gluster-dev
15:19 gbyte joined #gluster-dev
15:19 gbyte left #gluster-dev
15:40 jiffin joined #gluster-dev
15:42 cholcombe joined #gluster-dev
15:43 _Bryan_ joined #gluster-dev
16:08 soumya_ joined #gluster-dev
16:19 ashish joined #gluster-dev
16:55 rafi joined #gluster-dev
16:55 pranithk joined #gluster-dev
16:57 jiffin joined #gluster-dev
17:17 lpabon joined #gluster-dev
18:13 pousley joined #gluster-dev
18:13 RedW joined #gluster-dev
18:18 anoopcs joined #gluster-dev
18:26 cholcombe ndevos, nice notes on rpc https://github.com/nixpanic/gluster-wireshark/blob/master/RFC :)
18:52 ggarg joined #gluster-dev
18:54 wushudoin| joined #gluster-dev
18:59 wushudoin| joined #gluster-dev
19:01 shaunm_ joined #gluster-dev
19:17 cholcombe does null serialize in xdr to nothing or 0?
19:45 ndevos cholcombe: depends on the "null", I think, a string would be <length><data>, and <length> can be 0
19:46 cholcombe if the null is a iovec ?
19:46 ndevos cholcombe: it can hardly ever be nothing, otherwise it would be impossible to recreate the structure
19:48 ndevos the iovec would contain the payload, a read-call would have a null payload, I'd say
19:48 cholcombe ok
19:48 cholcombe just wondering
19:48 ndevos I'm mostly guessing, now, not looking at the code
19:54 dlambrig joined #gluster-dev
19:56 rafi joined #gluster-dev
20:10 shaunm_ joined #gluster-dev
20:29 jrm16020 joined #gluster-dev
20:46 an joined #gluster-dev
20:52 badone joined #gluster-dev
23:53 topshare joined #gluster-dev
23:54 shaunm_ joined #gluster-dev

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