On Wed, Nov 5, 2014 at 6:25 PM, Justin Clift jus...@gluster.org wrote:
Forwarding this, as it's possibly very relevant for the
GlusterD-2.0 stuff. It looks like the threading behaviour of Go
is very different to other languages (somewhat practical outline
given below).
(Start with the
On Sat, Oct 25, 2014 at 1:16 AM, Dennis Schafroth den...@schafroth.dk wrote:
Trying to test OSX towards a Linux (debian 7.7) server, but seems to get the
same error on both OS X and Linux when attempting to mount:
Running beta3 on both.
sudo mount -t glusterfs -o
On Sat, Oct 25, 2014 at 8:47 AM, Dennis Schafroth den...@schafroth.dk wrote:
Mounting a distributed replicated linux server sometimes shows duplicate
directory entries on mac.
Go ahead and open a bug :-)
--
Religious confuse piety with mere ritual, the virtuous confuse
regulation with
On Sat, Oct 25, 2014 at 11:13 AM, Harshavardhana
har...@harshavardhana.net wrote:
On Sat, Oct 25, 2014 at 8:47 AM, Dennis Schafroth den...@schafroth.dk wrote:
Mounting a distributed replicated linux server sometimes shows duplicate
directory entries on mac.
Go ahead and open a bug
Sure, I wasn't questioning that. I was expecting (guessing) that compiling
would need system fuse headers, but looking deeper I see that that seems not
to be the case.
Ah no not really, it exactly like how we build FUSE client for Linux -
there is no
external dependency other than the kernel
I have done a `brew fetch osxfuse` and I've got
/Library/Caches/Homebrew/osxfuse-2.7.2.yosemite.bottle.tar.gz.
??? Seems like I could just untar that in /usr/local/Cellar. Seem
reasonable?
We don't necessarily need OSXFUSE headers all, all we need is
/dev/fuse with the
kernel module loaded
On Tue, Sep 23, 2014 at 11:10 AM, Justin Clift jus...@gluster.org wrote:
On 23/09/2014, at 1:00 AM, Harshavardhana wrote:
Test Summary Report
---
./tests/bugs/bug-1109770.t (Wstat: 0 Tests: 19
Failed: 1)
Failed test: 13
Files=280, Tests=8149, 5589
On Mon, Sep 22, 2014 at 10:57 AM, Kaushal M kshlms...@gmail.com wrote:
I'm using Perl 5.20.1. I was wondering how no one caught a mistake in
the checker.
~kaushal
http://review.gluster.com/#/c/8811/ - here you go, please verify.
--
Religious confuse piety with mere ritual, the virtuous
On Fri, Sep 19, 2014 at 6:06 PM, Emmanuel Dreyfus m...@netbsd.org wrote:
Harshavardhana har...@harshavardhana.net wrote:
This is to bring in adherence to coding policy, prior to patch
submission for review.
- no tabs
- no whitespace
- indentation (linux style) etc.
What do we do whan
What do we do whan the code surrounding the change has tabs? Should we
adapt to the local style, or change everything?
It ignores changes in certain types of files, but .c and .h shouldn't have
tabs
So there are two possibilities
- Whitespace
- erroneous spaces between functions
-
/home/jenkins/root/workspace/rackspace-regression-2GB/xlators/cluster/dht/src/dht-common.c:
In function ‘dht_lookup_everywhere_done’:
/home/jenkins/root/workspace/rackspace-regression-2GB/xlators/cluster/dht/src/dht-common.c:1229:
warning: implicit declaration of function
Right, but that is a bug fixed in master but still present in
release-3.6, isn't it? Why not backport that change to release-3.6?
The patch will not apply cleanly, it requires previous changes, but
perhaps it is worth working on it?
That is left to the changeset owner, perhaps it will be
On Fri, Sep 5, 2014 at 9:34 AM, Emmanuel Dreyfus m...@netbsd.org wrote:
Hi
netbsd0.cloud.gluster.org now passes the POSIX test suite:
(...)
All tests successful.
Files=191, Tests=2069, 135 wallclock secs ( 1.03 usr 0.45 sys + 26.81 cusr
37.51 csys = 65.80 CPU)
Result: PASS
I have
There are couple of patches [1] submitted by me are resulting in hang. I
think these slaves were spawned to test the patch [1] and its dependencies.
If yes, they can be killed.
[1] http://review.gluster.com/#/c/8523/
One should be able to manually abort them in Jenkins.
--
Religious
In test_fstat():
msg=$(sh -c 'tail -f $PFX/dir/file --pid=$$ sleep 1 echo hooha
$PFX/dir/file sleep 1');
NetBSD does not have the --pid option. I propose this change, which
seems to obtain the same result with less complexity. Opinion?
-msg=$(sh -c 'tail -f $PFX/dir/file
Can you please a file a bug on this? we'll try to address it.
Sure submitting one right away.
https://bugzilla.redhat.com/show_bug.cgi?id=1133820 - here you go!
--
Religious confuse piety with mere ritual, the virtuous confuse
regulation with outcomes
On Fri, Aug 22, 2014 at 10:23 PM, Atin Mukherjee amukh...@redhat.com wrote:
IIRC, we were marking the verified as +1 in case of a known spurious
failure, can't we continue to do the same for the known spurious
failures just to unblock the patches getting merged till the problems
are resolved?
[3]: http://review.gluster.org/8352
Explored quite a few options here
- https://gist.github.com/harshavardhana/b5d3c69db5312b84022f - was
one approach which was worked by Brad Hubbard which makes used of
kernel dentry caches.
- For ext4 you have
# debugfs 'ncheck inode_num' device
- XFS provides
On Wed, Aug 20, 2014 at 1:55 PM, Ben England bengl...@redhat.com wrote:
What is update-link-count-parent POSIX xlator parameter for? Is it ever set
by anyone and why? It appears to be off by default. Didn't see this
documented, but the log message in posix.c says: update-link-count-parent
Again with the unpack issue with git
Build Failed
http://build.gluster.org/job/glusterfs-rpms/556/ : SUCCESS
http://build.gluster.org/job/glusterfs-rpms-el6/556/ : FAILURE
Receiving objects: 2% (2239/81742), 579.83 KiB | 854.00 KiB/s
Receiving objects: 3% (2453/81742), 579.83 KiB |
On Wed, Aug 20, 2014 at 2:49 PM, Justin Clift jus...@gluster.org wrote:
Any ideas?
Avati just fixed it
We had to run 'git gc' directly on the gerrit server - gerrit gc
failed (could be a gerrit bug).
Suspecting the EOF unpack errors to be gerrit bug as well, we should
be updating it to
Using git over ssh often fails, is it the same problem? I never
investigated, but using HTTP is reliable. I often have to run rfc.sh
three or four times before it works.
It seems like a 'gerrit' bug at the moment, its fixed with 'git gc' on
the server no corruption reported :-).
--
Not sure. We can figure this out by traversing up the softlinks for
directories. But for files there is no way to find the parent at the moment.
This email from Dan Mons puts some perspective on what actually people
expect - https://www.mail-archive.com/gluster-devel@gluster.org/msg00392.html
On Thu, Aug 14, 2014 at 11:12 AM, Joe Julian j...@julianfamily.org wrote:
Some people. Depends on use case. Dan's is pretty specific.
Those are majority of the users/customers.
--
Religious confuse piety with mere ritual, the virtuous confuse
regulation with outcomes
I made a patch that lets meta.t pass on NetBSD. Unfortunately, the
direct IO flag gets attached to the vnode and not to the file descriptor,
which means it is not possible to have a fd with direct IO and another
without.
But perhaps it is just good enough.
Can you point me to the change?
These are only NetBSD fixes:
http://ftp.espci.fr/shadow/manu/iodirect2.patch
netbsd0.cloud.gluster.org runs with them right now.
Looks like i can use this as a way to get direct io done for FUSE4BSD, this is
implemented exactly how i thought would be for FreeBSD. Let me see how much
effort
On Wed, Aug 13, 2014 at 9:45 PM, Anand Avati av...@gluster.org wrote:
Can you post a bt full?
Adding to that, is it a 32bit node? looks like it.
--
Religious confuse piety with mere ritual, the virtuous confuse
regulation with outcomes
___
This is a standard problem where there are split-brains in distributed
systems. For example even in git there are cases where it gives up asking
users to fix the file i.e. merge conflicts. If the user doesn't want
split-brains they should move to replica-3 and enable client-quorum. But if
the
I am working on tests/basic/meta.t on NetBSD
It fails because .meta/frames is empty, like all files in .meta. A quick
investigation in source code shows that the function responsible for
filling the code (frames_file_fill) is never called.
Same experience here.
It does work on OSX, but does
On Thu, Aug 7, 2014 at 1:35 AM, Ravishankar N ravishan...@redhat.com wrote:
Manual resolution of split-brains [1] has been a tedious task involving
understanding and modifying AFR's changelog extended attributes. To simplify
and to an extent automate this task, we are proposing a new CLI
Wait, directories *are* supposed to automatically heal from split-brain?
Guess I need to file a bug report. That doesn't happen. All the metadata and
gfid can be the same, but since the trusted.afr are both dirty, it'll stay
split-brain forever.
Conservative merge happens, but 'directories'
On Fri, Aug 8, 2014 at 12:53 PM, Joe Julian j...@julianfamily.org wrote:
Thinking about it more, I'd still rather have this functionality exposed at
the client through xattrs. For 5 years I've thought about this, and the more
I encounter split-brain, the more I think this is the needed
On Thu, Aug 7, 2014 at 2:41 AM, Niels de Vos nde...@redhat.com wrote:
On Wed, Aug 06, 2014 at 06:57:46PM -0700, Harshavardhana wrote:
FYI
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=192420
Nice! Is there a plan to propose glusterfs as a FreeBSD port too?
Yes, that is currently being
FreeBSD can vote as part of smoke tests, just like NetBSD.
Happy Porting!
--
Religious confuse piety with mere ritual, the virtuous confuse
regulation with outcomes
___
Gluster-devel mailing list
Gluster-devel@gluster.org
FYI
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=192420
Thanks
--
Religious confuse piety with mere ritual, the virtuous confuse
regulation with outcomes
___
Gluster-devel mailing list
Gluster-devel@gluster.org
On Thu, Jul 31, 2014 at 11:31 AM, Anders Blomdell
anders.blomd...@control.lth.se wrote:
During rsync of 35 files, memory consumption of glusterfs
rose to 12 GB (after approx 14 hours), I take it that this is a
bug I should try to track down?
Does it ever come down? what happens if you
On Mon, Jul 28, 2014 at 8:15 AM, Emmanuel Dreyfus m...@netbsd.org wrote:
On Mon, Jul 28, 2014 at 01:13:42PM +0100, Justin Clift wrote:
Just a heads up. The NetBSD and FreeBSD smoke tests have been enabled
for (at least) the release-3.6 and master branches.
release-3.6 did not build for
I've been meaning to ask (or suggest) this for some time
To me a smoke test should be a simple compile and run the executable(s) to
make sure they don't crash with the most basic configuration.
I think the posix tests that are in smoke.sh today should really be part of
the regression
This is The Right Way in my opinion. I think the current implementation
should not have been merged, but I do not track changes close enough to
had the opportunity to cast a -2 code review in time.
Note that a voting NetBSD build would have catched it. This changes restores
the build, we
Cmockery2 is a hard dependency before GlusterFS can be compiled in
upstream master now - we could make it conditional
and enable if necessary? since we know we do not have the cmockery2
packages available on all systems?
On Mon, Jul 21, 2014 at 10:16 AM, Luis Pabon lpa...@redhat.com wrote:
Niels
memory
leaks - perhaps it could be just FreeBSD thing!
On Wed, Jul 16, 2014 at 11:29 PM, Emmanuel Dreyfus m...@netbsd.org wrote:
On Wed, Jul 16, 2014 at 09:54:49PM -0700, Harshavardhana wrote:
In-fact this is true on Linux as well - there is smaller time window
observe the below output
process
#kill -USR1 pid of process
HTH,
Krish
- Original Message -
On Wed, Jul 16, 2014 at 11:32:06PM -0700, Harshavardhana wrote:
On a side note while looking into this issue - I uncovered a memory
leak too which after successful registration with glusterd, Self-heal
daemon
process
when the leak is observed would be really helpful.
thanks,
Krish
- Original Message -
Nope spoke too early, using poll() has no effect on the memory usage
on Linux, so actually back to FreeBSD.
On Thu, Jul 17, 2014 at 12:07 AM, Harshavardhana
har
Anyone?
On Mon, Jul 14, 2014 at 6:44 PM, Harshavardhana
har...@harshavardhana.net wrote:
http://review.gluster.org/#/c/8246/
Two important things it achieves
- Break-way from '/var/lib/glusterd' hard-coded previously,
instead rely on 'configure' value from 'localstatedir'
- Provide 's
is observed would be really helpful.
thanks,
Krish
- Original Message -
Nope spoke too early, using poll() has no effect on the memory usage
on Linux, so actually back to FreeBSD.
On Thu, Jul 17, 2014 at 12:07 AM, Harshavardhana
har...@harshavardhana.net wrote
is that this is related to
poll()/epoll() - i have to debug this further.
In-fact restarting 'gluster management daemon' fixes these issues
which is understandable :-)
On Wed, Jul 16, 2014 at 9:41 AM, Emmanuel Dreyfus m...@netbsd.org wrote:
Harshavardhana har...@harshavardhana.net wrote:
Its pretty much
I need reviews here to get these changes in for 3.6 , La FreeBSD -
http://review.gluster.com/#/c/8246/
On Tue, Jul 15, 2014 at 1:32 AM, Soumya Koduri skod...@redhat.com wrote:
Hi Vijay,
I suggest below patch to be merged -
http://review.gluster.org/#/c/7976/
Its not a critical one but fixes
Also for fuse_readlink bug from MacOSX testing -
http://review.gluster.com/#/c/8300/
On Tue, Jul 15, 2014 at 1:46 AM, Harshavardhana
har...@harshavardhana.net wrote:
I need reviews here to get these changes in for 3.6 , La FreeBSD -
http://review.gluster.com/#/c/8246/
On Tue, Jul 15, 2014
On Mon, Jul 14, 2014 at 9:31 AM, Anders Blomdell
anders.blomd...@control.lth.se wrote:
When submitting patches where there is an/some obvious person(s) to blame,
is it OK/desirable to request them as Code-Reviewers in gerrit?
Gist of adding Core-Reviewers is to find faults in oneself - not the
http://review.gluster.org/#/c/8246/
Two important things it achieves
- Break-way from '/var/lib/glusterd' hard-coded previously,
instead rely on 'configure' value from 'localstatedir'
- Provide 's/lib/db' as default working directory for gluster
management daemon for BSD and Darwin based
Do not know, they do not show up locally on my laptop, can you point
me to a build so that i can investigate?
On Thu, Jul 10, 2014 at 6:45 PM, Pranith Kumar Karampuri
pkara...@redhat.com wrote:
hi Harsha,
Know anything about the following warnings on latest master?
In file included from
On Thu, Jul 10, 2014 at 7:30 PM, Harshavardhana
har...@harshavardhana.net wrote:
Do not know, they do not show up locally on my laptop, can you point
me to a build so that i can investigate?
I think these are related to C99 standards, are you using clang? -
this must be xdrgen bug
I thought pkg-version in build-aux should have fixed this properly?
On Wed, Jul 9, 2014 at 1:33 PM, Justin Clift jus...@gluster.org wrote:
That v3.5qa2 tag name on master is annoying, due to the RPM
naming it causes when building on master.
Did we figure out a solution?
Maybe we should do a
On Thu, Jul 3, 2014 at 11:30 PM, Santosh Pradhan sprad...@redhat.com wrote:
Thanks guys for looking into this. I am just wondering how this passed the
regression before Niels could merged this in? Good part is test case needs
modification not code ;)
We need a single maintainer for test cases
There seems to be some bug in our regression testing code. Even though the
regression failed it gave the verdict as SUCCESS
http://build.gluster.org/job/rackspace-regression-2GB-triggered/97/consoleFull
This was fixed by Justin Clift recently
--
Religious confuse piety with mere ritual, the
Yeah, lets try this out. We can add the checkpatch.pl script to the
patch acceptance tests, and have an automatically triggered job that
runs it on patch submission. Should be pretty straightforward.
Let me work on the checkpatch script more to clean it up and make it
report properly for
Besides bandwidth limits, there also needs to be monitors on brick latency.
We don't want so many queued iops that operating performance is impacted.
AFAIK - rebalance and self-heal threads run in low-priority queue in
io-threads by default.
--
Religious confuse piety with mere ritual, the
Surprisingly why isn't this available in 'master' branch? isn't it
'master' first and then backport it to release-3.5?
On Sun, Jun 29, 2014 at 9:56 PM, Pranith Kumar Karampuri
pkara...@redhat.com wrote:
hi Louis,
It seems like 3.5.1 deb does not include the binary /usr/sbin/glfsheal
which
http://review.gluster.org/#/c/8181/ - posted a new change, wouldn't it
be worth to add this in smoke tests? rather than at ./rfc.sh ? - we
can provide a detailed summary - since we do not have 'commit/push'
style patch submission.
We can leverage our smoke tests, thoughts?
On Wed, Jun 25, 2014
Interesting - looks like all the sources have been moved? do we know where?
On Thu, Jun 12, 2014 at 10:48 PM, Justin Clift jus...@gluster.org wrote:
Hi Kaleb,
This just started showing up in rpm.t test output:
ERROR:
Agreed! +1
On Tue, Jun 10, 2014 at 7:51 PM, Pranith Kumar Karampuri
pkara...@redhat.com wrote:
hi,
I see that quota-anon-fd.t is causing too many spurious failures. I think
we should revert it and raise a bug so that it can be fixed and committed
again along with the fix.
Pranith
Emmanuel i sent a patch to disable building snapview on OSX, since it
was of no use
on 'darwin' .
You could see if snapshots and snapview would work on NetBSD (LVM
support), in such cases one
doesn't have to disable it for NetBSD.
d_off might be necessary internally, turning it off might not
It is always possible to translate structures, the question is whether
it is useful of not. d_off is the offset of this struct dirent within
the buffer for the whole directory returned by getdents(2) system call.
Since we glusterfs does not use getdents(2) but upper level
Do you reckon we should get that Mac Mini in the Westford
lab set up to automatically test Gluster builds each
night or something?
If so, we should probably take/claim ownership of it,
upgrade the memory in it, and (possibly) see if it can be
put in the DMZ.
Up to you guys, it would be
Here are the important locations in the XFS tree coming from 2.6.32 branch
STATIC int
xfs_set_acl(struct inode *inode, int type, struct posix_acl *acl)
{
struct xfs_inode *ip = XFS_I(inode);
unsigned char *ea_name;
int error;
if (S_ISLNK(inode-i_mode))
http://review.gluster.com/#/c/7823/ - the fix here
On Thu, May 22, 2014 at 1:41 PM, Harshavardhana
har...@harshavardhana.net wrote:
Here are the important locations in the XFS tree coming from 2.6.32 branch
STATIC int
xfs_set_acl(struct inode *inode, int type, struct posix_acl *acl
duplo# ls -li /mnt
total 4
4203451048 drwxr-xr-x 3 manu wheel 1024 May 16 14:08 manu
4203451048 drwxr-xr-x 3 manu wheel 1024 May 16 14:08 manu
3471060024 drwxrwxrwt 2 root wheel 1024 May 16 18:40 tmp
3471060024 drwxrwxrwt 2 root wheel 1024
You could try using that (then update your PATH to find /usr/local
stuff), too see if the problem goes away on OSX 10.8 with that.
If it works, that's a useful data point... ;)
Using gcc on OSX will mask a lot of apparent warning and cross
platform issues. clang values
conformance and Apps
68 matches
Mail list logo
