Re: [Gluster-users] Scaling Gluster
Use ucarp to share a virtual IP address between the Gluster servers, and mount using that IP address. Joshua Baker-LePain - Yes, step by step is here: http://www.misdivision.com/blog/setting-up-a-highly-available-storage-cluster- using-glusterfs-and-ucarp ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Scaling Gluster
I have though about this, It is the solution scalr offers, but the problem remains the same. What happen if the host which the dns mapped to is down? This solution solves load balancing but no high availability. On Mon, May 23, 2011 at 6:29 PM, Anthony J. Biacco < abia...@formatdynamics.com> wrote: > Make a DNS round-robin host entry that points to both IP addresses, then > use that hostname in the mount command > > > > -Tony > > --- > > Manager, IT Operations > > Format Dynamics, Inc. > > P: 303-228-7327 > > F: 303-228-7305 > > abia...@formatdynamics.com > > http://www.formatdynamics.com > > > > *From:* José Celano [mailto:jcel...@iqingenieros.com] > *Sent:* Monday, May 23, 2011 11:06 AM > *To:* Anthony J. Biacco > *Cc:* Joshua Baker-LePain; gluster-users@gluster.org > > *Subject:* Re: [Gluster-users] Scaling Gluster > > > > Hi Tony: > > On Mon, May 23, 2011 at 4:55 PM, Anthony J. Biacco < > abia...@formatdynamics.com> wrote: > > I think i read somewhere the server in the mount string is only used to > retrieve the cluster config. So as long as that server is up at time of > mount, you're fine. > > > Yes, It is in the offcial documentation. > I am trying to avoid that. My final objetive is high availability. > > > If the server goes down after mount, it doesnt matter as the cluster config > on the mounting servers knows about all gluster servers. > > > Yes, there is no problem after mounting becasuse I use replication. The > problem is in new mountig processes > after the seleted "master gluster server" crashes. > > > Somebody correct me if i'm wrong. > Personally, i copy the cluster config to all my mounting servers and use > that file in the mount command instead of a gluster server hostname. > > > Yes. That is ok, but I am finding an automatic way in case I make an > auto-scalable Gluster deployment. > If something change in Gluster like the number of servers (I could add more > servers later) then I would have > to copy the new volumen info to my Apache Rol Amazon Image, and this can > not be done automatically. > Other way could be doing a script to load that file from S3 but I would > have to copy the new volume info file > to S3 every time It changes, and I don´t know how this can be done because > I don´t know how monitoring > and catch Gluster volumes changes. > > > -Tony > > > > > > -Original Message- > From: Joshua Baker-LePain > Sent: May 23, 2011 9:16 AM > To: José Celano > Cc: gluster-users@gluster.org > Subject: Re: [Gluster-users] Scaling Gluster > > On Thu, 19 May 2011 at 12:40am, José Celano wrote > > > I have 2 replicated servers and several Apache servers that mount the > same > > volume. > > When I mount a Gluster Volumen from one of the Apache servers > > > > mount.glusterfs > > > ip-XX-XXX-XXX-XXX.eu-west-1.compute.internal:/dab578f4-06fa-4584-ac1d-c6ef14e6d0cf > > /mnt/glusterfs > > how could I obtain high availability if that server die? > > Use ucarp to share a virtual IP address between the Gluster servers, and > mount using that IP address. > > -- > Joshua Baker-LePain > QB3 Shared Cluster Sysadmin > UCSF > > > ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Scaling Gluster
Other solution I have been thinking in is making a script that before mounting the volume in the client side makes a previous test (perhaps ping or even the same mount command) and if this test fails It just tries it with the next Gluster server in the list. The problem here is to obtain the dynamic list of all gluster severs. I am using sclar.netand they do that job for me. There is a directory in every instance where they automatically put all active server IPs for every rol. But I am finding a simpler solution because I thought it is strange thatGluster don´t have a way of doing this internally. Perhaps It could be done adding a gluster mount command with 2 o 3 "master servers" location parameters and that it chooses internally the first is not down (like DNS). On Mon, May 23, 2011 at 5:15 PM, Anthony J. Biacco < abia...@formatdynamics.com> wrote: > I don't expect and haven't made many volume changes, so the maintenance > involved transferring files every so often versus maintenance of another > puzzle piece that can fail is a no brainer for me. > But I'm sure many others have more complicated and ever changing setups > than me. > > -Tony > --- > Manager, IT Operations > Format Dynamics, Inc. > P: 303-228-7327 > F: 303-228-7305 > abia...@formatdynamics.com > http://www.formatdynamics.com > > > > -Original Message- > > From: Joshua Baker-LePain [mailto:jl...@duke.edu] > > Sent: Monday, May 23, 2011 10:02 AM > > To: Anthony J. Biacco > > Cc: José Celano; gluster-users@gluster.org > > Subject: RE: [Gluster-users] Scaling Gluster > > > > On Mon, 23 May 2011 at 9:55am, Anthony J. Biacco wrote > > > > > I think i read somewhere the server in the mount string is only used to > > retrieve the cluster config. So as long as that server is up at time of > mount, > > you're fine. > > > If the server goes down after mount, it doesnt matter as the cluster > config > > on the mounting servers knows about all gluster servers. > > > Somebody correct me if i'm wrong. > > > > That's correct -- it's the mount-time failure that ucarp can help avoid. > > > > > Personally, i copy the cluster config to all my mounting servers and > use > > > that file in the mount command instead of a gluster server hostname. > > > > But then you lose the ability to easily and automatically distribute > > configuration changes made via the 'gluster' command on the servers. I'm > > not saying that can't be worked around, but everything is a tradeoff. > > > > -- > > Joshua Baker-LePain > > QB3 Shared Cluster Sysadmin > > UCSF > > ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Scaling Gluster
Hi Tony: On Mon, May 23, 2011 at 4:55 PM, Anthony J. Biacco < abia...@formatdynamics.com> wrote: > I think i read somewhere the server in the mount string is only used to > retrieve the cluster config. So as long as that server is up at time of > mount, you're fine. > Yes, It is in the offcial documentation. I am trying to avoid that. My final objetive is high availability. > If the server goes down after mount, it doesnt matter as the cluster config > on the mounting servers knows about all gluster servers. > Yes, there is no problem after mounting becasuse I use replication. The problem is in new mountig processes after the seleted "master gluster server" crashes. > Somebody correct me if i'm wrong. > Personally, i copy the cluster config to all my mounting servers and use > that file in the mount command instead of a gluster server hostname. > Yes. That is ok, but I am finding an automatic way in case I make an auto-scalable Gluster deployment. If something change in Gluster like the number of servers (I could add more servers later) then I would have to copy the new volumen info to my Apache Rol Amazon Image, and this can not be done automatically. Other way could be doing a script to load that file from S3 but I would have to copy the new volume info file to S3 every time It changes, and I don´t know how this can be done because I don´t know how monitoring and catch Gluster volumes changes. > -Tony > > > > > -Original Message- > From: Joshua Baker-LePain > Sent: May 23, 2011 9:16 AM > To: José Celano > Cc: gluster-users@gluster.org > Subject: Re: [Gluster-users] Scaling Gluster > > On Thu, 19 May 2011 at 12:40am, José Celano wrote > > > I have 2 replicated servers and several Apache servers that mount the > same > > volume. > > When I mount a Gluster Volumen from one of the Apache servers > > > > mount.glusterfs > > > ip-XX-XXX-XXX-XXX.eu-west-1.compute.internal:/dab578f4-06fa-4584-ac1d-c6ef14e6d0cf > > /mnt/glusterfs > > how could I obtain high availability if that server die? > > Use ucarp to share a virtual IP address between the Gluster servers, and > mount using that IP address. > > -- > Joshua Baker-LePain > QB3 Shared Cluster Sysadmin > UCSF > > ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Scaling Gluster
On Mon, May 23, 2011 at 4:16 PM, Joshua Baker-LePain wrote: > On Thu, 19 May 2011 at 12:40am, José Celano wrote > > > I have 2 replicated servers and several Apache servers that mount the same >> volume. >> When I mount a Gluster Volumen from one of the Apache servers >> >> mount.glusterfs >> >> ip-XX-XXX-XXX-XXX.eu-west-1.compute.internal:/dab578f4-06fa-4584-ac1d-c6ef14e6d0cf >> /mnt/glusterfs >> how could I obtain high availability if that server die? >> > > Use ucarp to share a virtual IP address between the Gluster servers, and > mount using that IP address. > It seems that solution don´t work because you have to use fixed IP for virtual IP and Amazon Elastic IP isn´t exactly a fixed IP. I have only found this information about this issue: and http://gluster.org/pipermail/gluster-users/2011-March/006759.html and http://gluster.com/community/documentation/index.php/GlusterFS_1.3_P2P_Cluster_with_Auto_Healing See also: http://support.scalr.net/discussions/questions/746-scaling-gluster > > -- > Joshua Baker-LePain > QB3 Shared Cluster Sysadmin > UCSF ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Scaling Gluster
I did a quick test of the round-robin with 2 servers (1 server down) and it worked ok with 10 consecutive mount tries. -Tony --- Manager, IT Operations Format Dynamics, Inc. P: 303-228-7327 F: 303-228-7305 abia...@formatdynamics.com http://www.formatdynamics.com <http://www.formatdynamics.com/> From: José Celano [mailto:jcel...@iqingenieros.com] Sent: Monday, May 23, 2011 11:40 AM To: Anthony J. Biacco Cc: Joshua Baker-LePain; gluster-users@gluster.org Subject: Re: [Gluster-users] Scaling Gluster I have though about this, It is the solution scalr offers, but the problem remains the same. What happen if the host which the dns mapped to is down? This solution solves load balancing but no high availability. On Mon, May 23, 2011 at 6:29 PM, Anthony J. Biacco wrote: Make a DNS round-robin host entry that points to both IP addresses, then use that hostname in the mount command -Tony --- Manager, IT Operations Format Dynamics, Inc. P: 303-228-7327 F: 303-228-7305 abia...@formatdynamics.com http://www.formatdynamics.com <http://www.formatdynamics.com/> From: José Celano [mailto:jcel...@iqingenieros.com] Sent: Monday, May 23, 2011 11:06 AM To: Anthony J. Biacco Cc: Joshua Baker-LePain; gluster-users@gluster.org Subject: Re: [Gluster-users] Scaling Gluster Hi Tony: On Mon, May 23, 2011 at 4:55 PM, Anthony J. Biacco wrote: I think i read somewhere the server in the mount string is only used to retrieve the cluster config. So as long as that server is up at time of mount, you're fine. Yes, It is in the offcial documentation. I am trying to avoid that. My final objetive is high availability. If the server goes down after mount, it doesnt matter as the cluster config on the mounting servers knows about all gluster servers. Yes, there is no problem after mounting becasuse I use replication. The problem is in new mountig processes after the seleted "master gluster server" crashes. Somebody correct me if i'm wrong. Personally, i copy the cluster config to all my mounting servers and use that file in the mount command instead of a gluster server hostname. Yes. That is ok, but I am finding an automatic way in case I make an auto-scalable Gluster deployment. If something change in Gluster like the number of servers (I could add more servers later) then I would have to copy the new volumen info to my Apache Rol Amazon Image, and this can not be done automatically. Other way could be doing a script to load that file from S3 but I would have to copy the new volume info file to S3 every time It changes, and I don´t know how this can be done because I don´t know how monitoring and catch Gluster volumes changes. -Tony -Original Message- From: Joshua Baker-LePain Sent: May 23, 2011 9:16 AM To: José Celano Cc: gluster-users@gluster.org Subject: Re: [Gluster-users] Scaling Gluster On Thu, 19 May 2011 at 12:40am, José Celano wrote > I have 2 replicated servers and several Apache servers that mount the same > volume. > When I mount a Gluster Volumen from one of the Apache servers > > mount.glusterfs > ip-XX-XXX-XXX-XXX.eu-west-1.compute.internal:/dab578f4-06fa-4584-ac1d-c6ef14e6d0cf > /mnt/glusterfs > how could I obtain high availability if that server die? Use ucarp to share a virtual IP address between the Gluster servers, and mount using that IP address. -- Joshua Baker-LePain QB3 Shared Cluster Sysadmin UCSF ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Scaling Gluster
I always thought it retried for another IP after a timeout, though I haven't test it. -Tony --- Manager, IT Operations Format Dynamics, Inc. P: 303-228-7327 F: 303-228-7305 abia...@formatdynamics.com http://www.formatdynamics.com <http://www.formatdynamics.com/> From: José Celano [mailto:jcel...@iqingenieros.com] Sent: Monday, May 23, 2011 11:40 AM To: Anthony J. Biacco Cc: Joshua Baker-LePain; gluster-users@gluster.org Subject: Re: [Gluster-users] Scaling Gluster I have though about this, It is the solution scalr offers, but the problem remains the same. What happen if the host which the dns mapped to is down? This solution solves load balancing but no high availability. On Mon, May 23, 2011 at 6:29 PM, Anthony J. Biacco wrote: Make a DNS round-robin host entry that points to both IP addresses, then use that hostname in the mount command -Tony --- Manager, IT Operations Format Dynamics, Inc. P: 303-228-7327 F: 303-228-7305 abia...@formatdynamics.com http://www.formatdynamics.com <http://www.formatdynamics.com/> From: José Celano [mailto:jcel...@iqingenieros.com] Sent: Monday, May 23, 2011 11:06 AM To: Anthony J. Biacco Cc: Joshua Baker-LePain; gluster-users@gluster.org Subject: Re: [Gluster-users] Scaling Gluster Hi Tony: On Mon, May 23, 2011 at 4:55 PM, Anthony J. Biacco wrote: I think i read somewhere the server in the mount string is only used to retrieve the cluster config. So as long as that server is up at time of mount, you're fine. Yes, It is in the offcial documentation. I am trying to avoid that. My final objetive is high availability. If the server goes down after mount, it doesnt matter as the cluster config on the mounting servers knows about all gluster servers. Yes, there is no problem after mounting becasuse I use replication. The problem is in new mountig processes after the seleted "master gluster server" crashes. Somebody correct me if i'm wrong. Personally, i copy the cluster config to all my mounting servers and use that file in the mount command instead of a gluster server hostname. Yes. That is ok, but I am finding an automatic way in case I make an auto-scalable Gluster deployment. If something change in Gluster like the number of servers (I could add more servers later) then I would have to copy the new volumen info to my Apache Rol Amazon Image, and this can not be done automatically. Other way could be doing a script to load that file from S3 but I would have to copy the new volume info file to S3 every time It changes, and I don´t know how this can be done because I don´t know how monitoring and catch Gluster volumes changes. -Tony -Original Message- From: Joshua Baker-LePain Sent: May 23, 2011 9:16 AM To: José Celano Cc: gluster-users@gluster.org Subject: Re: [Gluster-users] Scaling Gluster On Thu, 19 May 2011 at 12:40am, José Celano wrote > I have 2 replicated servers and several Apache servers that mount the same > volume. > When I mount a Gluster Volumen from one of the Apache servers > > mount.glusterfs > ip-XX-XXX-XXX-XXX.eu-west-1.compute.internal:/dab578f4-06fa-4584-ac1d-c6ef14e6d0cf > /mnt/glusterfs > how could I obtain high availability if that server die? Use ucarp to share a virtual IP address between the Gluster servers, and mount using that IP address. -- Joshua Baker-LePain QB3 Shared Cluster Sysadmin UCSF ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Scaling Gluster
Make a DNS round-robin host entry that points to both IP addresses, then use that hostname in the mount command -Tony --- Manager, IT Operations Format Dynamics, Inc. P: 303-228-7327 F: 303-228-7305 abia...@formatdynamics.com http://www.formatdynamics.com <http://www.formatdynamics.com/> From: José Celano [mailto:jcel...@iqingenieros.com] Sent: Monday, May 23, 2011 11:06 AM To: Anthony J. Biacco Cc: Joshua Baker-LePain; gluster-users@gluster.org Subject: Re: [Gluster-users] Scaling Gluster Hi Tony: On Mon, May 23, 2011 at 4:55 PM, Anthony J. Biacco wrote: I think i read somewhere the server in the mount string is only used to retrieve the cluster config. So as long as that server is up at time of mount, you're fine. Yes, It is in the offcial documentation. I am trying to avoid that. My final objetive is high availability. If the server goes down after mount, it doesnt matter as the cluster config on the mounting servers knows about all gluster servers. Yes, there is no problem after mounting becasuse I use replication. The problem is in new mountig processes after the seleted "master gluster server" crashes. Somebody correct me if i'm wrong. Personally, i copy the cluster config to all my mounting servers and use that file in the mount command instead of a gluster server hostname. Yes. That is ok, but I am finding an automatic way in case I make an auto-scalable Gluster deployment. If something change in Gluster like the number of servers (I could add more servers later) then I would have to copy the new volumen info to my Apache Rol Amazon Image, and this can not be done automatically. Other way could be doing a script to load that file from S3 but I would have to copy the new volume info file to S3 every time It changes, and I don´t know how this can be done because I don´t know how monitoring and catch Gluster volumes changes. -Tony -Original Message- From: Joshua Baker-LePain Sent: May 23, 2011 9:16 AM To: José Celano Cc: gluster-users@gluster.org Subject: Re: [Gluster-users] Scaling Gluster On Thu, 19 May 2011 at 12:40am, José Celano wrote > I have 2 replicated servers and several Apache servers that mount the same > volume. > When I mount a Gluster Volumen from one of the Apache servers > > mount.glusterfs > ip-XX-XXX-XXX-XXX.eu-west-1.compute.internal:/dab578f4-06fa-4584-ac1d-c6ef14e6d0cf > /mnt/glusterfs > how could I obtain high availability if that server die? Use ucarp to share a virtual IP address between the Gluster servers, and mount using that IP address. -- Joshua Baker-LePain QB3 Shared Cluster Sysadmin UCSF ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Scaling Gluster
I don't expect and haven't made many volume changes, so the maintenance involved transferring files every so often versus maintenance of another puzzle piece that can fail is a no brainer for me. But I'm sure many others have more complicated and ever changing setups than me. -Tony --- Manager, IT Operations Format Dynamics, Inc. P: 303-228-7327 F: 303-228-7305 abia...@formatdynamics.com http://www.formatdynamics.com > -Original Message- > From: Joshua Baker-LePain [mailto:jl...@duke.edu] > Sent: Monday, May 23, 2011 10:02 AM > To: Anthony J. Biacco > Cc: José Celano; gluster-users@gluster.org > Subject: RE: [Gluster-users] Scaling Gluster > > On Mon, 23 May 2011 at 9:55am, Anthony J. Biacco wrote > > > I think i read somewhere the server in the mount string is only used to > retrieve the cluster config. So as long as that server is up at time of mount, > you're fine. > > If the server goes down after mount, it doesnt matter as the cluster config > on the mounting servers knows about all gluster servers. > > Somebody correct me if i'm wrong. > > That's correct -- it's the mount-time failure that ucarp can help avoid. > > > Personally, i copy the cluster config to all my mounting servers and use > > that file in the mount command instead of a gluster server hostname. > > But then you lose the ability to easily and automatically distribute > configuration changes made via the 'gluster' command on the servers. I'm > not saying that can't be worked around, but everything is a tradeoff. > > -- > Joshua Baker-LePain > QB3 Shared Cluster Sysadmin > UCSF ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Scaling Gluster
On Mon, 23 May 2011 at 9:55am, Anthony J. Biacco wrote I think i read somewhere the server in the mount string is only used to retrieve the cluster config. So as long as that server is up at time of mount, you're fine. If the server goes down after mount, it doesnt matter as the cluster config on the mounting servers knows about all gluster servers. Somebody correct me if i'm wrong. That's correct -- it's the mount-time failure that ucarp can help avoid. Personally, i copy the cluster config to all my mounting servers and use that file in the mount command instead of a gluster server hostname. But then you lose the ability to easily and automatically distribute configuration changes made via the 'gluster' command on the servers. I'm not saying that can't be worked around, but everything is a tradeoff. -- Joshua Baker-LePain QB3 Shared Cluster Sysadmin UCSF ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Scaling Gluster
I think i read somewhere the server in the mount string is only used to retrieve the cluster config. So as long as that server is up at time of mount, you're fine. If the server goes down after mount, it doesnt matter as the cluster config on the mounting servers knows about all gluster servers. Somebody correct me if i'm wrong. Personally, i copy the cluster config to all my mounting servers and use that file in the mount command instead of a gluster server hostname. -Tony -Original Message- From: Joshua Baker-LePain Sent: May 23, 2011 9:16 AM To: José Celano Cc: gluster-users@gluster.org Subject: Re: [Gluster-users] Scaling Gluster On Thu, 19 May 2011 at 12:40am, José Celano wrote > I have 2 replicated servers and several Apache servers that mount the same > volume. > When I mount a Gluster Volumen from one of the Apache servers > > mount.glusterfs > ip-XX-XXX-XXX-XXX.eu-west-1.compute.internal:/dab578f4-06fa-4584-ac1d-c6ef14e6d0cf > /mnt/glusterfs > how could I obtain high availability if that server die? Use ucarp to share a virtual IP address between the Gluster servers, and mount using that IP address. -- Joshua Baker-LePain QB3 Shared Cluster Sysadmin UCSF ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Scaling Gluster
On Thu, 19 May 2011 at 12:40am, José Celano wrote I have 2 replicated servers and several Apache servers that mount the same volume. When I mount a Gluster Volumen from one of the Apache servers mount.glusterfs ip-XX-XXX-XXX-XXX.eu-west-1.compute.internal:/dab578f4-06fa-4584-ac1d-c6ef14e6d0cf /mnt/glusterfs how could I obtain high availability if that server die? Use ucarp to share a virtual IP address between the Gluster servers, and mount using that IP address. -- Joshua Baker-LePain QB3 Shared Cluster Sysadmin UCSF___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Scaling Gluster
> "José" == José Celano writes: José> When I mount a Gluster Volumen from one of the Apache José> servers José> mount.glusterfs José> ip-XX-XXX-XXX-XXX.eu-west-1.compute.internal:/ José> dab578f4-06fa-4584-ac1d-c6ef14e6d0cf /mnt/glusterfs José> how could I obtain high availability if that server die? You get that automatically. The gluster client will use any available bricks; obviously you need a mirrored setup to make that work. -- All the best, Berend de Boer -- Awesome Drupal hosting: https://www.xplainhosting.com/ ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users