Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster-dev, 2016-06-06

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

All times shown according to UTC.

Time Nick Message
01:47 ilbot3 joined #gluster-dev
01:47 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/
01:56 josferna joined #gluster-dev
02:25 josferna joined #gluster-dev
02:45 kshlm joined #gluster-dev
02:46 josferna joined #gluster-dev
03:17 nigelb review.gluster.org is down for quick maintenance.
03:17 nigelb The login issues should be sorted once we're back on line.
03:21 nishanth joined #gluster-dev
03:37 nigelb review.gluster.org is back online, the login issues should be sorted now.
03:46 hagarth nigelb: thanks, was about to check with you a few minutes back about r.g.o :)
03:49 nigelb hagarth: :)
03:51 itisravi joined #gluster-dev
03:54 nbalacha joined #gluster-dev
03:56 overclk joined #gluster-dev
04:01 atinm joined #gluster-dev
04:10 ppai joined #gluster-dev
04:15 nigelb hagarth: Your username has likely changed after the fix up.
04:17 hagarth nigelb: checked right now.. does seem to have stayed the same
04:18 ppai nigelb, so I'm able to login now, if any issue crops up later in the day, I'll let you know. Thanks.
04:18 ppai nigelb++
04:18 glusterbot ppai: nigelb's karma is now 6
04:18 nigelb ppai: hurray! :)
04:31 nigelb hagarth: is your account name vijay or vbellur?
04:31 nigelb vbellur is what it should be now.
04:35 kdhananjay joined #gluster-dev
04:36 kotreshhr joined #gluster-dev
04:37 josferna joined #gluster-dev
04:38 aspandey joined #gluster-dev
04:46 hagarth nigelb: it is vbellur now
04:46 prasanth joined #gluster-dev
04:46 hagarth nigelb: I thought it was vbellur previously too? if not, please overlook my ignorance ;)
04:47 nigelb hagarth: You may have been pushing to git.gluster.org as vijay. I've been looking at the sshd error logs to help people who have auth problems.
04:47 nigelb I saw an auth failure from you, so I thought I'd let you know what's wrong :)
04:47 nigelb (The actual error message is extremely unhelpful, it tells you that your authentication failed, not that your username is wrong)
04:49 kdhananjay joined #gluster-dev
04:50 hagarth nigelb: right, i just changed vijay to vbellur in my git config now
04:50 nigelb :)
04:50 hagarth nigelb: thank you for noticing that! :)
04:51 nigelb kdhananjay: You're having problems pushing to git.gluster.org as well, correct?
04:51 kdhananjay joined #gluster-dev
04:53 kdhananjay nigelb: git pull results in the following error: "Permission denied (publickey). fatal: Could not read from remote repository. Please make sure you have the correct access rights and the repository exists."
04:53 nigelb kdhananjay: I know. I just hit send on an email to you about that.
04:54 * kdhananjay is going through it.
04:55 shubhendu joined #gluster-dev
05:08 anoopcs joined #gluster-dev
05:11 nigelb kdhananjay: any luck in solving that?
05:13 kdhananjay nigelb: i responded to your mail.
05:20 gem joined #gluster-dev
05:23 karthik___ joined #gluster-dev
05:28 ndarshan joined #gluster-dev
05:36 poornimag joined #gluster-dev
05:43 mchangir joined #gluster-dev
05:46 Bhaskarakiran joined #gluster-dev
05:48 Apeksha joined #gluster-dev
05:48 hgowtham joined #gluster-dev
05:49 jiffin joined #gluster-dev
05:49 hchiramm joined #gluster-dev
06:08 ppai joined #gluster-dev
06:18 ashiq joined #gluster-dev
06:19 rafi joined #gluster-dev
06:32 aravindavk joined #gluster-dev
06:36 spalai joined #gluster-dev
06:45 Manikandan joined #gluster-dev
06:49 pur__ joined #gluster-dev
06:59 pk joined #gluster-dev
06:59 pranithk1 joined #gluster-dev
06:59 pk jiffin: http://review.gluster.org/#/c/4591/
07:00 Guest71295 joined #gluster-dev
07:03 jiffin pk: thanks
07:16 pkalever joined #gluster-dev
07:22 rastar joined #gluster-dev
07:25 Saravanakmr joined #gluster-dev
07:26 spalai1 joined #gluster-dev
07:30 ndevos_ nigelb:
07:30 ndevos nigelb++ I can login on review.gluster.org again \o/
07:30 glusterbot ndevos: nigelb's karma is now 7
07:31 nigelb ndevos: yay! :)
07:47 Jules- joined #gluster-dev
07:49 josferna joined #gluster-dev
08:01 post-factum http://termbin.com/px54
08:01 post-factum awesome
08:01 post-factum glustershd VIRT is 72.8G
08:01 post-factum !!!
08:03 karthik___ joined #gluster-dev
08:03 post-factum this is just 3rd node in pool that keeps volumes metadata, nothing more
08:04 sakshi joined #gluster-dev
08:05 post-factum should i care?
08:17 rraja joined #gluster-dev
08:36 pranithk1 joined #gluster-dev
08:36 pk joined #gluster-dev
08:38 aspandey joined #gluster-dev
08:39 itisravi_ joined #gluster-dev
08:40 kdhananjay joined #gluster-dev
08:41 atalur joined #gluster-dev
08:50 anoop_ joined #gluster-dev
08:52 spalai joined #gluster-dev
08:52 anoop_ joined #gluster-dev
08:53 anoop_ joined #gluster-dev
08:57 kshlm post-factum, are you using encrypted transports by any chance? I've seen huge virt sizes for glusterd with encryption enabled.
08:58 post-factum kshlm: nope, no encryption here
08:58 anoop_ joined #gluster-dev
08:58 post-factum kshlm: i've sent detailed description to ML
08:58 anoop_ left #gluster-dev
08:58 kshlm Ah, okay.
09:05 kshlm Is multi-threaded shd in 3.7.11?
09:06 kshlm pranithk1, ^
09:20 nigelb kshlm: Ha, the issue of gerrit accidentally aggregating into verified tag is exactly what I ran into on that patch, I think :)
09:22 kshlm nigelb, Yup. This might be a bug with the jenkins plugin.
09:23 kshlm The jobs are set to be launched from events on the different configured servers, and for different events.
09:23 kshlm But sometimes, it aggregates the smoke jobs under the wrong server.
09:24 nigelb Yeah :|
09:32 raghu joined #gluster-dev
09:47 atinm joined #gluster-dev
09:47 mchangir joined #gluster-dev
10:13 kkeithley is something going on with git.gluster.org?  I just tried a `git pull` of glusterfs and got permission denied.
10:14 Guest71295 kkeithley, some of us had our gerrit usernames changed. I had to update by git remote to reflect new ssh username
10:16 nigelb kkeithley: Your gerrit username is now the same as your github username.
10:16 nigelb So you need to edit your git remote to reflect that.
10:16 kkeithley ???  my gerrit/github username have always been the same.
10:16 kkeithley but I use ssh to clone.
10:17 spalai1 joined #gluster-dev
10:17 nigelb kkeithley: On ssh, I bet you used a username that's not your github username.
10:17 atalur joined #gluster-dev
10:17 kkeithley yes
10:18 nigelb The fact that it worked is basically a stroke of luck.
10:18 aspandey_ joined #gluster-dev
10:18 nigelb That should have been the same as your github username too.
10:19 shubhendu joined #gluster-dev
10:20 itisravi joined #gluster-dev
10:25 ndarshan joined #gluster-dev
10:27 nishanth joined #gluster-dev
10:30 kdhananjay joined #gluster-dev
10:32 poornimag kdhananjay, do you have upstream bug for the VM corruption that you saw in case of replace brick? If so can you please share the bug id?
10:45 mchangir joined #gluster-dev
10:52 atalur joined #gluster-dev
10:54 pk joined #gluster-dev
10:54 pranithk1 joined #gluster-dev
11:05 karthik___ joined #gluster-dev
11:06 shubhendu joined #gluster-dev
11:08 ndarshan joined #gluster-dev
11:11 nishanth joined #gluster-dev
11:13 atalur joined #gluster-dev
11:25 itisravi pranithk1: review requested for http://review.gluster.org/#/c/14642/
11:27 anoopcs kkeithley, I could see glusterfs-coreutils component under Fedora for bugzilla. I think we agreed on to have coreutils as a component for GlusterFS.
11:29 pranithk1 itisravi: Could you get a review from one more person before I do the review?
11:29 itisravi pranithk1: sure I'll ask.
11:29 itisravi atalur: aspandey_ kdhananjay ^
11:32 aspandey joined #gluster-dev
11:35 aspandey nigelb, regression is failing however https://build.gluster.org/job/rackspace-r​egression-2GB-triggered/19466/consoleFull shows nothing has failed. What can we do to get correct result?
11:40 nigelb I'll look in a few minutes
11:47 ppai joined #gluster-dev
11:49 kkeithley anoopcs: srsly?  sigh.  Here's the ticket https://engineering.redhat.com/r​t/Ticket/Display.html?id=406332
11:51 anoopcs kkeithley, Oh..Ok. No replies yet. Let's wait.
11:52 anoopcs btw, scratch builds completed successfully on f22, f23, f24.
11:52 atinm joined #gluster-dev
11:54 kkeithley glusterfs-coreutils is there as a Fedora component. Which is not what what was asked for.
11:58 kkeithley unless Fedora/glusterfs-coreutils happened automagically as a result of the package
11:59 anoopcs kkeithley, I think that's automatically created.
11:59 kkeithley as I think about it, yes, that's probably what happened.
11:59 anoopcs for every new package that gets approved.
11:59 kkeithley so Fedora/glusterfs-coreutils is for dist-git and packaging bugs and that sort of thing.
12:00 nigelb aspandey: This is the same issue kaushal and I were talking about earlier.
12:00 nigelb Every so often, it seems to happen.
12:00 kkeithley and GlusterFS/coreutils will be for everything else
12:07 kotreshhr left #gluster-dev
12:11 ppai joined #gluster-dev
12:21 penguinRaider joined #gluster-dev
12:25 spalai1 joined #gluster-dev
12:51 ira joined #gluster-dev
13:03 nbalacha joined #gluster-dev
13:32 jiffin joined #gluster-dev
13:34 josferna joined #gluster-dev
13:34 shyam joined #gluster-dev
13:44 luizcpg joined #gluster-dev
13:50 pur__ joined #gluster-dev
13:52 gem joined #gluster-dev
13:53 josferna joined #gluster-dev
13:54 pranithk1 joined #gluster-dev
14:11 pranithk1 joined #gluster-dev
14:31 pranithk1 joined #gluster-dev
15:00 pranithk1 joined #gluster-dev
15:03 hchiramm joined #gluster-dev
15:05 pranithk1 ndevos: kkeithley: hey I am seeing that writes in nfs are coming with O_SYNC flag, is that expected?
15:05 pranithk1 ndevos: kkeithley: This is with nfsv3 not ganesha
15:07 wushudoin joined #gluster-dev
15:08 hagarth pranithk1: what are the mount options for the client?
15:09 kkeithley doesn't seem right. I don't think there have been any changes in the nfs server xlator though
15:09 wushudoin joined #gluster-dev
15:10 pranithk1 hagarth: normal mount -t nfs -o vers=3 `hostname`:/volname /mnt/point
15:10 pranithk1 kkeithley: hagarth: checking with v3.5.0 code just to see if it is the same or not
15:11 spalai joined #gluster-dev
15:13 pranithk1 hagarth: kkeithley: Even with 3.5.0, I get O_SYNC in the flags: Breakpoint 1, posix_writev (frame=0x7f09394c93a8, this=0xdc0f00, fd=0xdfbfcc,
15:13 pranithk1 vector=0xdbca50, count=1, offset=0, flags=1052672, iobref=0xdbc200, xdata=0x7f0938ec4768)
15:13 pranithk1 at posix.c:2543
15:13 pranithk1 2543{
15:17 shyam left #gluster-dev
15:17 hagarth pranithk1: please paste the output from mount -t nfs on the client
15:19 pranithk1 hagarth: one sec
15:20 pranithk1 hagarth: localhost.localdomain:/r2 on /mnt/nfs type nfs (rw,relatime,vers=3,rsize=1048576,wsize=1048576,n​amlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=​sys,mountaddr=127.0.0.1,mountvers=3,mountport=384​65,mountproto=tcp,local_lock=none,addr=127.0.0.1)
15:21 pranithk1 hagarth: I didn't give all these options though
15:21 hagarth pranithk1: these are the defaults .. wanted to check what the default options are
15:21 aspandey joined #gluster-dev
15:22 hagarth pranithk1: it does look like the client is sending out writes as stable
15:22 hagarth pranithk1: check nfs3_write_resume()
15:22 hagarth we add O_SYNC if the client marks the writetype as stable/file_sync
15:24 pranithk1 hagarth: It is...
15:25 pranithk1 hagarth: How to disable it?
15:25 hagarth pranithk1: what workload are you running on the client?
15:27 pranithk1 hagarth: I want to change some code around not modifying fd->flags in nfs which is creating problems for o-direct fds in sharding. Krutika did that already at http://review.gluster.org/14191
15:27 pranithk1 hagarth: But I want to change it again to not touch fd->flags ever...
15:28 pranithk1 hagarth: Like in http://paste.fedoraproject.org/375404/46522686/
15:29 pranithk1 hagarth: when I was testing this, I always see writetype to STABLE whether I use oflag=sync or not in nfs. So not able to test it
15:34 hagarth pranithk1: agree with Jeff's comments in Krutika's patch
15:34 pranithk1 hagarth: https://build.gluster.org/job/rackspace-r​egression-2GB-triggered/19474/consoleFull
15:35 pranithk1 hagarth: No dude, write just needs flags in write fop to be present
15:35 pranithk1 hagarth: which it is anyway sending
15:37 hagarth pranithk1: what are we checking in posix? write's flags or fd->flags?
15:37 hagarth or both?
15:37 pranithk1 hagarth: flags
15:38 hagarth pranithk1: why is fd->flags being set?
15:41 pranithk1 hagarth: There was no need
15:43 pranithk1 hagarth: Gave reply to comment
15:43 pranithk1 hagarth: The code is misunderstood
15:43 hagarth pranithk1: ok
15:44 shyam joined #gluster-dev
15:44 pranithk1 hagarth: So any tips?
15:44 pranithk1 hagarth: For disabling stable writes?
15:47 hagarth pranithk1: an easy way would be to modify the server code
15:47 hagarth and test
15:48 pranithk1 hagarth: meaning? cs->writetype?
15:48 hagarth yeah .. just don't set O_SYNC in nfs server for testing
15:50 pranithk1 hagarth: okay
16:03 shaunm joined #gluster-dev
16:06 pranithk1 hagarth: done with testing :-)
16:08 hagarth pranithk1: cool :)
16:15 pranithk1 hagarth: Even when https://build.gluster.org/job/rackspace-r​egression-2GB-triggered/19474/consoleFull is success, it posted failure :-/
16:16 hagarth pranithk1: probably related to the email that poornima sent?
16:16 pranithk1 hagarth: ignore
17:06 pranithk1 hagarth: what is the mail you said poornimag sent?
17:24 hagarth pranithk1: check on gluster-infra
17:25 pranithk1 hagarth: I am not on gluster-infra :-)
17:26 hagarth pranithk1: the archives are public
17:26 pranithk1 hagarth: cool
17:39 atalur joined #gluster-dev
17:48 shubhendu joined #gluster-dev
17:59 pranithk1 hagarth: Are you in office?
18:01 hagarth pranithk1: no, am wfh today
18:01 pranithk1 hagarth: Is there a way to reach jdarcy other than mail?
18:01 pranithk1 hagarth: I want to talk to him about -2 on the patch
18:02 hagarth pranithk1: sending an email is the easiest I guess
18:02 pranithk1 hagarth: okay
18:13 pranithk1 nigelb: Are you online? Quite A few regressions are failing with build errors... Doesn't seem like gluster problem to me at least. Any clue?
18:19 pranithk1 hagarth: :-/ wonder how to get around these problems :-(
18:19 gem joined #gluster-dev
18:30 hagarth pranithk1: nigelb and kaushal are our best bets atm
18:30 hagarth pranithk1: do send a note on gluster-infra
18:31 pranithk1 hagarth: both of them are not online :-). Is there something that you know and you can guide me to solve these problems?
18:33 hagarth pranithk1: I am not aware of recent changes in infra .. can you point me to a specific problem that you are trying to address?
18:33 pranithk1 hagarth: regressions are failing at random places
18:34 pranithk1 hagarth: The links it posts to gerrit are old builds. Let me give and example
18:34 post-factum pranithk1: saw your email about memleak, will try to cherry-pick that patch and watch again
18:34 post-factum pranithk1: thanks for pointing to it
18:36 hagarth pranithk1: drop a note to amye also .. she might be able to help as well
18:36 pranithk1 hagarth: http://review.gluster.org/#/c/14649, search for rackspace-regression-2GB, you will see old build ids
18:37 hagarth pranithk1: no idea .. seeing this for the first time. maybe there's some compatibility break between the upgraded gerrit and old jenkins?
18:38 pranithk1 hagarth: Why was the upgrade done?
18:38 hagarth pranithk1: the above is a crude guess. looking into gerrit & jenkins logs would help.
18:38 hagarth pranithk1: to move away from h2 database, which is difficult to manage in the long run.
18:39 pranithk1 hagarth: hmm...
18:42 hagarth i don't think i anymore have access to the logs too
18:47 pranithk1 hagarth: Sent mails to people to look into.
18:49 hagarth pranithk1: ok, cool
18:54 gem joined #gluster-dev
18:55 jiffin joined #gluster-dev
19:03 luizcpg joined #gluster-dev
19:11 pranithk1 post-factum: np!
19:20 nishanth joined #gluster-dev
19:44 shyam left #gluster-dev
19:46 shyam joined #gluster-dev
20:26 wushudoin joined #gluster-dev
22:15 luizcpg joined #gluster-dev

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