Re: [Gluster-users] sync - sshd at different port.
Am 01.10.2013 00:59, schrieb Michael Post: Hello together, Hi Michael, i setup geo-replication. In my environment my ssh server use a different port. Where can i setup the different port, so gsync use the different port? As a quick workaround, I'd suggest to use a tunnel which will connect/"map" remote-ip:remote-nonstandard-port with/to local-ip:standard-port (cf. netcat, stunnel). Kind regards, Markus ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] Using gluster with ipv6
Am 30.09.2013 16:24, schrieb Mickaël Villers: Hi, Hi, I'm starting to use gluster version debian/unstable 3.4.0-4 but I need use ipv6 in my network, but just can't setup glusterfs for. I tried configuring IPv6 or DNS (only resolving at IPv6), got errors in both case. I also need IPv6 support (AFAIK it's not possible to use a server whose IPv4 address might change which would require repeated DNS lookups[*])--however, as a quick workaround, using stunnel comes to mind, which lets you specify a local IPv4 address and in turn will connect to the real IPv6 address through a tunnel. Alternatively, ssh/netcat might both be worth a look (the latter should allow to turn off encryption if two-factor authentication is not needed and you want to get rid of the "encryption overhead"). [*] that scenario can be addressed by re-establishing the aforementioned tunnel whenever the IP address changes Kind regards, Markus ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] Advice for building samba-glusterfs-vfs
Dan, I had the same trouble yesterday. As it happens, I created a doc to help script installs of future nodes. I did not snip out the portions that apply to what you're looking for, but the script below works for me. The biggest issue was that some modules were apparently installed in lib and the build process was looking for them in lib64. In any event, the vfs module builds, installs and runs cleanly after this. ** The big win for me was finding this command, that allowed me to figure out where it was looking for modules in the wrong lib directory. ldd /usr/local/samba/lib/vfs/glusterfs.so I'm sure there is any easier way to to do this :)... Cheers, Dave -- #!/bin/bash yum groupinstall "Development Tools" yum install git openssl-devel wget yum install libtalloc libtdb # set up gluster cd /usr/src && git clone https://github.com/gluster/glusterfs.git cd /usr/src/glusterfs && ./autogen.sh && ./configure && make make install # set up samba 3.6.9 cd /usr/src && wget http://ftp.samba.org/pub/samba/stable/samba-3.6.9.tar.gz&&; tar -zxvf samba-3.6.9.tar.gz cd /usr/src/samba-3.6.9/source3 && ./configure && make make install ln -s /usr/local/samba/lib/libwbclient.so.0 /usr/lib64/libwbclient.so.0 # then install the RPM samba version yum install samba # set up vfs_glusterfs cd /usr/src && git clone git:// forge.gluster.org/samba-glusterfs/samba-glusterfs-vfs.git ln -s /usr/local/include/glusterfs /usr/include/glusterfs cd /usr/src/samba-vfs/glusterfs ./configure --with-samba-source=/usr/src/samba-3.6.9/source3 ln -s /usr/local/samba/lib/vfs/glusterfs.so /usr/lib64/samba/vfs/glusterfs.so # link the other modules ln -s /usr/local/lib/libgfapi.so /usr/lib64/ ln -s /usr/local/lib/libgfapi.so.0 /usr/lib64/ ln -s /usr/local/lib/libgfapi.la /usr/lib64/ ln -s /usr/local/lib/libglusterfs.la /usr/lib64/ ln -s /usr/local/lib/libglusterfs.so /usr/lib64/ ln -s /usr/local/lib/libglusterfs.so.0 /usr/lib64/ ln -s /usr/local/lib/libglusterfs.so.0.0.0 /usr/lib64/ EOF On Tue, Oct 1, 2013 at 10:38 PM, Dan Mons wrote: > Hi folks, > > I've got CentOS6.4 with Samba 3.6.9 installed from the standard CentOS > repos via yum. I also have GlusterFS 3.4.0 GA installed from RPMs > direct from gluster.org. > > I'm trying to build the glusterfs VFS module for Samba to take > advantage of libgfapi for our Windows users, and migrate them off the > current Samba-on-FUSE setup we have currently. > > I've downloaded the appropriate source trees for all projects > (GlusterFS from gluster.org, Samba from the matching CentOS6 SRPM, and > samba-glusterfs-vfs from the git repo), but am facing troubles early > on just finding appropriate headers. > > [root@bne-gback000 samba-glusterfs-vfs]# find > /usr/local/src/glusterfs-3.4.0 -type f -name glfs.h > /usr/local/src/glusterfs-3.4.0/api/src/glfs.h > > [root@bne-gback000 samba-glusterfs-vfs]# ./configure > --with-glusterfs=/usr/local/src/glusterfs-3.4.0 > *snip* > checking api/glfs.h usability... no > checking api/glfs.h presence... no > checking for api/glfs.h... no > Cannot find api/glfs.h. Please specify --with-glusterfs=dir if necessary > > If I install glusterfs-api-devel-3.4.0-8.el6.x86_64.rpm, I need to > copy /usr/include/glusterfs/api/glfs.h to /usr/include for it to be > found (even using --with-glusterfs= doesn't work), and then I get > further errors about not being able to link to glfs_init: > > [root@bne-gback000 samba-glusterfs-vfs]# rpm -ivh > /tmp/glusterfs-api-devel-3.4.0-8.el6.x86_64.rpm > [root@bne-gback000 samba-glusterfs-vfs]# cp > /usr/include/glusterfs/api/glfs.h /usr/include/ > [root@bne-gback000 samba-glusterfs-vfs]# ./configure > *snip* > checking api/glfs.h usability... yes > checking api/glfs.h presence... yes > checking for api/glfs.h... yes > checking for glfs_init... no > Cannot link to gfapi (glfs_init). Please specify --with-glusterfs=dir > if necessary > > If anyone can point me in the right direction, that would be greatly > appreciated. > > Cheers, > > -Dan > ___ > Gluster-users mailing list > Gluster-users@gluster.org > http://supercolony.gluster.org/mailman/listinfo/gluster-users > ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] Managing volumes is unreliable
On 10/02/2013 06:57 AM, Heggland, Christian wrote: Unsure if its recommended but you could try and do the following if the set command stops adding / removing settings: 1. Stop the volume 2. Edit the config file for that volume on each node ( /var/lib/glusterd/vols ) 3. Start the volume again. I've had to do this in order to update the access list since the set command stopped working. Definitely not recommended. Any changes you make that way are likely to be overwritten the next time the standard management mechanisms are used (including at system startup). There is apparently a bug in the glusterd connection code. We need to concentrate on fixing that, and then see where things stand. ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] Managing volumes is unreliable
On 2 October 2013 12:57, Heggland, Christian wrote: > Unsure if its recommended but you could try and do the following if the set > command stops adding / removing settings: > > 1. Stop the volume > 2. Edit the config file for that volume on each node ( /var/lib/glusterd/vols > ) > 3. Start the volume again. It still requires stopping and starting volume which may fail. regards -- Maciej Gałkiewicz Shelly Cloud Sp. z o. o., Sysadmin http://shellycloud.com/, mac...@shellycloud.com KRS: 440358 REGON: 101504426 ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] Managing volumes is unreliable
Unsure if its recommended but you could try and do the following if the set command stops adding / removing settings: 1. Stop the volume 2. Edit the config file for that volume on each node ( /var/lib/glusterd/vols ) 3. Start the volume again. I've had to do this in order to update the access list since the set command stopped working. Best Regards Christian -Original Message- From: gluster-users-boun...@gluster.org [mailto:gluster-users-boun...@gluster.org] On Behalf Of Maciej Galkiewicz Sent: 2. oktober 2013 11:35 To: Gluster General Discussion List Subject: [Gluster-users] Managing volumes is unreliable Hi I am using glusterfs with following scenario: Multiple untrusted groups of clients (hosts) need access to shared storage. For each group a separate glusterfs volume is created with authentication based on IP addresses. Groups and clients are dynamically created/removed so there are a lot of operations on volumes like start, stop, info, set. Unfortunately managing volumes with those command is quite unreliable. I am affected by https://bugzilla.redhat.com/show_bug.cgi?id=1002422. Is there any other suitable for my scenario way of using glusterfs which could reduce operations on volumes or somehow workaround this issue? regards -- Maciej Gałkiewicz Shelly Cloud Sp. z o. o., Sysadmin http://shellycloud.com/, mac...@shellycloud.com KRS: 440358 REGON: 101504426 ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] Shutting down a GlusterFS server.
Hi Hoggins, Thanks. I have set gluster services to start during startup. I believe the self-heal would play the most important role of taking care of the changes in the dead-peer once it comes back up. Any comments or suggestions for the same ? Thanks & Regards, Bobby Jacob -Original Message- From: gluster-users-boun...@gluster.org [mailto:gluster-users-boun...@gluster.org] On Behalf Of Hoggins! Sent: Wednesday, October 02, 2013 12:46 PM To: gluster-users@gluster.org Subject: Re: [Gluster-users] Shutting down a GlusterFS server. Hello Bobby, I experienced problems shutting down Gluster servers, because the startup scripts (SystemD) did not have dependencies for system shutdown that would stop the Gluster services before anything else. My problems were also related to the fact that my nodes are connected to each other through IPSec links, and during shutdown, the links were cut before the Gluster service was properly shut down. The other node was waiting for the well-known 42 seconds before considering the peer dead, hence not responding to any file request, and hanging all client systems. Not cool. So the best practice is simply to end the Gluster services before shutting down the server. Everything runs as smooth as it should be, and the clients don't notice any downtime. Hope this helps. Hoggins! Le 02/10/2013 11:36, Bobby Jacob a écrit : > > Hi, > > > > I have a 2-node replica volume running with GlusterFS 3.3.2 on Centos > 6.4. I want to shut down one of the gluster servers for maintenance. > Any best practice that is to be followed while turning off a server in > terms of services etc. Or can I just shut down the server. ? > ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
Re: [Gluster-users] Shutting down a GlusterFS server.
Hello Bobby, I experienced problems shutting down Gluster servers, because the startup scripts (SystemD) did not have dependencies for system shutdown that would stop the Gluster services before anything else. My problems were also related to the fact that my nodes are connected to each other through IPSec links, and during shutdown, the links were cut before the Gluster service was properly shut down. The other node was waiting for the well-known 42 seconds before considering the peer dead, hence not responding to any file request, and hanging all client systems. Not cool. So the best practice is simply to end the Gluster services before shutting down the server. Everything runs as smooth as it should be, and the clients don't notice any downtime. Hope this helps. Hoggins! Le 02/10/2013 11:36, Bobby Jacob a écrit : > > Hi, > > > > I have a 2-node replica volume running with GlusterFS 3.3.2 on Centos > 6.4. I want to shut down one of the gluster servers for maintenance. > Any best practice that is to be followed while turning off a server in > terms of services etc. Or can I just shut down the server. ? > ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
[Gluster-users] Shutting down a GlusterFS server.
Hi, I have a 2-node replica volume running with GlusterFS 3.3.2 on Centos 6.4. I want to shut down one of the gluster servers for maintenance. Any best practice that is to be followed while turning off a server in terms of services etc. Or can I just shut down the server. ? Thanks & Regards, Bobby Jacob ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
[Gluster-users] Managing volumes is unreliable
Hi I am using glusterfs with following scenario: Multiple untrusted groups of clients (hosts) need access to shared storage. For each group a separate glusterfs volume is created with authentication based on IP addresses. Groups and clients are dynamically created/removed so there are a lot of operations on volumes like start, stop, info, set. Unfortunately managing volumes with those command is quite unreliable. I am affected by https://bugzilla.redhat.com/show_bug.cgi?id=1002422. Is there any other suitable for my scenario way of using glusterfs which could reduce operations on volumes or somehow workaround this issue? regards -- Maciej Gałkiewicz Shelly Cloud Sp. z o. o., Sysadmin http://shellycloud.com/, mac...@shellycloud.com KRS: 440358 REGON: 101504426 ___ Gluster-users mailing list Gluster-users@gluster.org http://supercolony.gluster.org/mailman/listinfo/gluster-users
