a_pirania at poczta.onet.pl
2009-Jan-07 11:15 UTC
[Gluster-users] glusterfs alternative ? :P
I know that this is not the appropriate place :). You know someone can alternative to gluserfs ?:) -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20090107/63b68a0d/attachment.html>
it depends what you''re trying to accomplish with your filesystem. there are many alternatives each of which are suitable for specific needs. if you want a general purpose solution that is feature rich and solves a wide variety of problems, then gluster is really the best choice but depending on your specific nees, there may be better alternatives out there. At 03:15 AM 1/7/2009, a_pirania at poczta.onet.pl wrote:>I know that this is not the appropriate place :). You know someone >can alternative to gluserfs ?:) >_______________________________________________ >Gluster-users mailing list >Gluster-users at gluster.org >http://zresearch.com/cgi-bin/mailman/listinfo/gluster-users
it depends what you're trying to accomplish with your filesystem. there are many alternatives each of which are suitable for specific needs. if you want a general purpose solution that is feature rich and solves a wide variety of problems, then gluster is really the best choice but depending on your specific nees, there may be better alternatives out there. At 03:15 AM 1/7/2009, a_pirania at poczta.onet.pl wrote:>I know that this is not the appropriate place :). You know someone >can alternative to gluserfs ?:) >_______________________________________________ >Gluster-users mailing list >Gluster-users at gluster.org >http://zresearch.com/cgi-bin/mailman/listinfo/gluster-users
a_pirania at poczta.onet.pl
2009-Jan-07 11:30 UTC
[Gluster-users] glusterfs alternative ? :P
"Keith Freedman" <freedman at FreeFormIT.com> napisa?(a): > it depends what you're trying to accomplish with your filesystem. > > there are many alternatives each of which are suitable for specific needs. > if you want a general purpose solution that is feature rich and > solves a wide variety of problems, then gluster is really the best > choice but depending on your specific nees, there may be better > alternatives out there. > > I need a redundant system files in the event of failure of the first. glusterfs is a lot of errors
At 03:30 AM 1/7/2009, a_pirania at poczta.onet.pl wrote:> >"Keith Freedman" <freedman at FreeFormIT.com> napisa?(a): > > it depends what you''re trying to accomplish with your filesystem. > > > > there are many alternatives each of which are suitable for specific needs. > > if you want a general purpose solution that is feature rich and > > solves a wide variety of problems, then gluster is really the best > > choice but depending on your specific nees, there may be better > > alternatives out there. > > > > >I need a redundant system files in the event of >failure of the first. glusterfs is a lot of errorswell, gluster 2.0 isn''t really out yet, it''s in beta, so it''s likely to be a little error prone. but it''s the only one I''ve found which has good live failover at all. you could explore coda which is agony to install, or ocfs2 which is fine if you have shared storage. there''s cloudstore (I think that''s what they call it now), but if I remember correctly, it''s not a fuse based filesystem, so and is only accessible via an API, which makes it pretty useless unless you''re developing a specific application. google "clustered storage with failover" and you''ll get a list to explore.
a_pirania at poczta.onet.pl
2009-Jan-07 13:15 UTC
[Gluster-users] glusterfs alternative ? :P
"Keith Freedman" <freedman at FreeFormIT.com> napisa?(a): > At 04:28 AM 1/7/2009, a_pirania at poczta.onet.pl wrote: > >What errors are you getting? :-) We surely want to fix any errors. > > > >Krishna > > > > > >You read thread "AFR self-heal problem" :) > > > make sure you're on the latest patch-level, I'm pretty sure those > problems have been resolved. > I ran into that same problem and it was fixed the same day I reported it. > > Keith > Next thread "AFR not working ?" there is a problem in glusterfs 1.4.0rc3 built on Dec 17 2008 15:34:25 Repository revision: glusterfs--mainline--3.0--patch-777
At 05:15 AM 1/7/2009, a_pirania at poczta.onet.pl wrote:>"Keith Freedman" <freedman at FreeFormIT.com> napisa?(a): > > At 04:28 AM 1/7/2009, a_pirania at poczta.onet.pl wrote: > > >What errors are you getting? :-) We surely want to fix any errors. > > > > > >Krishna > > > > > > > > >You read thread "AFR self-heal problem" :) > > > > > > make sure you''re on the latest patch-level, I''m pretty sure those > > problems have been resolved. > > I ran into that same problem and it was fixed the same day I reported it. > > > > Keith > > > >Next thread "AFR not working ?" there is a >problem in glusterfs 1.4.0rc3 built on Dec 17 2008 15:34:25 >Repository revision: glusterfs--mainline--3.0--patch-777well, rc3/patch 777 are very old. I''m running patch-824 (from tla) and the problem you''re concerned with is not happening in my setup.
a_pirania at poczta.onet.pl
2009-Jan-07 13:36 UTC
[Gluster-users] glusterfs alternative ? :P
"Keith Freedman" <freedman at FreeFormIT.com> napisa?(a): > At 05:15 AM 1/7/2009, a_pirania at poczta.onet.pl wrote: > >"Keith Freedman" <freedman at FreeFormIT.com> napisa?(a): > > > At 04:28 AM 1/7/2009, a_pirania at poczta.onet.pl wrote: > > > >What errors are you getting? :-) We surely want to fix any errors. > > > > > > > >Krishna > > > > > > > > > > > >You read thread "AFR self-heal problem" :) > > > > > > > > > make sure you're on the latest patch-level, I'm pretty sure those > > > problems have been resolved. > > > I ran into that same problem and it was fixed the same day I reported it. > > > > > > Keith > > > > > > >Next thread "AFR not working ?" there is a > >problem in glusterfs 1.4.0rc3 built on Dec 17 2008 15:34:25 > >Repository revision: glusterfs--mainline--3.0--patch-777 > > well, rc3/patch 777 are very old. > I'm running patch-824 (from tla) and the problem > you're concerned with is not happening in my setup. > > Ok I check it
Dnia 7 stycznia 2009 22:37 Keith Freedman <freedman at FreeFormIT.com> napisa?(a):> At 07:11 AM 1/7/2009, you wrote: > > > > >a_pirania at poczta.onet.pl napisa?(a): > > > > > > > > > "Keith Freedman" napisa?(a): > > > > At 05:15 AM 1/7/2009, a_pirania at poczta.onet.pl wrote: > > > > >"Keith Freedman" napisa?(a): > > > > > > At 04:28 AM 1/7/2009, a_pirania at poczta.onet.pl wrote: > > > > > > >What errors are you getting? :-) We > > surely want to fix any errors. > > > > > > > > > > > > > >Krishna > > > > > > > > > > > > > > > > > > > > >You read thread "AFR self-heal problem" :) > > > > > > > > > > > > > > > > > > make sure you're on the latest patch-level, I'm pretty sure those > > > > > > problems have been resolved. > > > > > > I ran into that same problem and it > > was fixed the same day I reported it. > > > > > > > > > > > > Keith > > > > > > > > > > > > > > > >Next thread "AFR not working ?" there is a > > > > >problem in glusterfs 1.4.0rc3 built on Dec 17 2008 15:34:25 > > > > >Repository revision: glusterfs--mainline--3.0--patch-777 > > > > > > > > well, rc3/patch 777 are very old. > > > > I'm running patch-824 (from tla) and the problem > > > > you're concerned with is not happening in my setup. > > > > > > > > > > > Ok > > > I check it > > > > > >ehhhhhh bug! > > > >2009-01-07 16:03:19 E > >[afr-self-heal-data.c:778:afr_sh_data_fix] afr: > >Unable to resolve conflicting data of > >/blogdev/production/files/rupy1.jpg. Please > >resolve manually by deleting the file > >/blogdev/production/files/rupy1.jpg from all but > >the preferred subvolume. Please consider 'option favorite-child <>' > > > > > > > >noc-test-2:/mnt/glusterfs# cat ./blogdev/production/files/rupy1.jpg | head -1 > >cat: ./blogdev/production/files/rupy1.jpg: Input/output error > > > >glusterfs 1.4.0rc7 built on Jan 7 2009 15:00:10 > >Repository revision: glusterfs--mainline--3.0--patch-814 > > > not bug.. you need to read these error messages. > auto-heal can't figure out which file is the > newest one, so instead of just picking and doing > something bad, it's letting you fix it. > you can fix it by either deleting the file from > all but the preferred volume OR you can add: > option favorite-child > to the gluster config. > > if you choose to add option favorite-child then > it will always pick from that brick whenever there's an un resolvable conflict. >Why is there an error message in the log file, when only one server is running ? There is no conflict to resolve in such case
Hi, The two interesting open source parallel distributed fault-tolerant file systems I have found and tried are GlusterFS and Ceph. Perhaps Lustre will be interesting for me in the future if they fix fault-tolerance without special shared block storage on the servers. (Or have they?) Mirroring the data two or more servers is good enough for me, although I would love some RAID-6/RAID-Z -like redundancy. Ceph is found here <http://ceph.newdream.net/>. I did some benchmarking for Ceph a few months ago <http://www.update.uu.se/~jerker/ceph/>. On Ceph I got around 65 MByte/s write bandwidth on one node (using "dd") and around 120 MByte/s aggregate for the whole cluster (7 nodes). It was around a year ago the last time I configured and ran GlusterFS on the machines, but I do plan to do some better benchmarking for both file systems in the not so distant future. --jerker On Wed, 7 Jan 2009, a_pirania at poczta.onet.pl wrote:> I know that this is not the appropriate place :). You know someone can > alternative to gluserfs ?:) >
what I''m noticing is, with AFR/HA in gluster it seems that in the general case, 1.3 seems overall faster from a user experience perspective. 2.0 is a bit more robust and with that robustness seems to come a slight performance problem. when you do your benchmarking do so with some very large directories and some small ones, and some which change rapidly and some which change infrequently. it''ll be interesting to see how that impacts things.. In my empirical observations, a large directory has a large impact in 2.0. it''s entirely possible most of the performance related items I''ve notices are resolved, I''ll need more time to see how things behave. At 07:17 AM 1/8/2009, Jerker Nyberg wrote:>Hi, > >The two interesting open source parallel distributed fault-tolerant file >systems I have found and tried are GlusterFS and Ceph. Perhaps Lustre will >be interesting for me in the future if they fix fault-tolerance without >special shared block storage on the servers. (Or have they?) Mirroring the >data two or more servers is good enough for me, although I would love some >RAID-6/RAID-Z -like redundancy. > >Ceph is found here <http://ceph.newdream.net/>. I did some benchmarking >for Ceph a few months ago <http://www.update.uu.se/~jerker/ceph/>. On Ceph >I got around 65 MByte/s write bandwidth on one node (using "dd") and >around 120 MByte/s aggregate for the whole cluster (7 nodes). > >It was around a year ago the last time I configured and ran GlusterFS on >the machines, but I do plan to do some better benchmarking for both file >systems in the not so distant future. > >--jerker > > >On Wed, 7 Jan 2009, a_pirania at poczta.onet.pl wrote: > > > I know that this is not the appropriate place :). You know someone can > > alternative to gluserfs ?:) > > > >_______________________________________________ >Gluster-users mailing list >Gluster-users at gluster.org >http://zresearch.com/cgi-bin/mailman/listinfo/gluster-users
On Mon, 12 Jan 2009, Keith Freedman wrote:> The advantages gluster has over all the other cluster filesystems out > there are:The concept "cluster file system" may sometimes be confusing since it may mean both a shared disk file system (like OCFS, GFS) and a distributed file system intended for cluster usage (like GlusterFS, Lustre, PVFS and Ceph).> 1) you can use commodity hardware without having to build additional > infrastructure (you don't need a SAN, just the same computers > connected to the same network will do the trick)As far as I know all parallel distributed file systems (PVFS, Lustre, GlusterFS, Ceph, etc) may be run on commodity hardware. However, to get redundancy or failover for Lustre and PVFS you need shared block storage. The shared disk ones (OCFS, GFS) do of course not support this.> 2) replica's (ha/afr, whatever) needn't be physically connected (or > connected by anything but a network connection. While it might be a > tad slow, it will work exactly the same over a wide area as within a > data center.Both Ceph and GlusterFS share this advantage which in my opinion make them both very interesting. (And I am ignoring solutions with DRBD for Lustre and PVFS.) Regards, Jerker Nyberg.
On Mon, 12 Jan 2009, Keith Freedman wrote:> (starting with CODA--whatever happened to that?).InterMezzo was started as part of the Coda file system project at Carnegie Mellon University. Later the developers moved on to Lustre. At least according to Wikipedia. http://en.wikipedia.org/wiki/InterMezzo_(file_system) Regards, Jerker Nyberg.