Post reply

Cluster Quorum - Node and Disk

  • October 31, 2012 at 1:18 am

    #283844

    Hi Guys

    Please help me understand something the Cluster Quorum and specifically Node and Disk Majority

    I understand that if you have an even amount on nodes on your cluster, it's recommended to implement Node and Disk Majority. To avoid the split-brain effect and other undesirables.

    My question is, if your disk that you have specified as your Quorum disk goes down, does it take the entire cluster down?

    Thanks

  • October 31, 2012 at 2:47 am

    #1554509

    SQLSACT (10/31/2012)

    Hi Guys

    Please help me understand something the Cluster Quorum and specifically Node and Disk Majority

    I understand that if you have an even amount on nodes on your cluster, it's recommended to implement Node and Disk Majority. To avoid the split-brain effect and other undesirables.

    My question is, if your disk that you have specified as your Quorum disk goes down, does it take the entire cluster down?

    Thanks

    Not necessarily, it depends on what other votes are around at the time of failure. More on the new quorum models can be found at this link.

    Also, the link above states

    Technet online

    Can sustain failures of half the nodes (rounding up) if the disk witness remains online. For example, a six node cluster in which the disk witness is online could sustain three node failures.

    Can sustain failures of half the nodes (rounding up) minus one if the disk witness goes offline or fails. For example, a six node cluster with a failed disk witness could sustain two (3-1=2) node failures.

    -----------------------------------------------------------------------------------------------------------

    "Ya can't make an omelette without breaking just a few eggs" 😉

  • October 31, 2012 at 3:51 am

    #1554522

    Perry Whittle (10/31/2012)

    SQLSACT (10/31/2012)

    Hi Guys

    Please help me understand something the Cluster Quorum and specifically Node and Disk Majority

    I understand that if you have an even amount on nodes on your cluster, it's recommended to implement Node and Disk Majority. To avoid the split-brain effect and other undesirables.

    My question is, if your disk that you have specified as your Quorum disk goes down, does it take the entire cluster down?

    Thanks

    Not necessarily, it depends on what other votes are around at the time of failure. More on the new quorum models can be found at this link.

    Also, the link above states

    Technet online

    Can sustain failures of half the nodes (rounding up) if the disk witness remains online. For example, a six node cluster in which the disk witness is online could sustain three node failures.

    Can sustain failures of half the nodes (rounding up) minus one if the disk witness goes offline or fails. For example, a six node cluster with a failed disk witness could sustain two (3-1=2) node failures.

    Thanks

    In my case, I have a 2 node 2008 cluster with Node and Disk Majority

    I took down the Quorum drive and the entire cluster went down

    Normal?

  • October 31, 2012 at 4:09 am

    #1554524

    SQLSACT (10/31/2012)

    Perry Whittle (10/31/2012)

    SQLSACT (10/31/2012)

    Hi Guys

    Please help me understand something the Cluster Quorum and specifically Node and Disk Majority

    I understand that if you have an even amount on nodes on your cluster, it's recommended to implement Node and Disk Majority. To avoid the split-brain effect and other undesirables.

    My question is, if your disk that you have specified as your Quorum disk goes down, does it take the entire cluster down?

    Thanks

    Not necessarily, it depends on what other votes are around at the time of failure. More on the new quorum models can be found at this link.

    Also, the link above states

    Technet online

    Can sustain failures of half the nodes (rounding up) if the disk witness remains online. For example, a six node cluster in which the disk witness is online could sustain three node failures.

    Can sustain failures of half the nodes (rounding up) minus one if the disk witness goes offline or fails. For example, a six node cluster with a failed disk witness could sustain two (3-1=2) node failures.

    Thanks

    In my case, I have a 2 node 2008 cluster with Node and Disk Majority

    I took down the Quorum drive and the entire cluster went down

    Normal?

    No, as 2 out of 3 votes would have still been available.

    Having said that, in your config, removal of the disk witness would not sustain any further node failures.

    Re run the cluster validation report and check for errors, also check the event and cluster logs for issues too

    -----------------------------------------------------------------------------------------------------------

    "Ya can't make an omelette without breaking just a few eggs" 😉

  • October 31, 2012 at 4:35 am

    #1554533

    Perry Whittle (10/31/2012)

    SQLSACT (10/31/2012)

    Perry Whittle (10/31/2012)

    SQLSACT (10/31/2012)

    Hi Guys

    Please help me understand something the Cluster Quorum and specifically Node and Disk Majority

    I understand that if you have an even amount on nodes on your cluster, it's recommended to implement Node and Disk Majority. To avoid the split-brain effect and other undesirables.

    My question is, if your disk that you have specified as your Quorum disk goes down, does it take the entire cluster down?

    Thanks

    Not necessarily, it depends on what other votes are around at the time of failure. More on the new quorum models can be found at this link.

    Also, the link above states

    Technet online

    Can sustain failures of half the nodes (rounding up) if the disk witness remains online. For example, a six node cluster in which the disk witness is online could sustain three node failures.

    Can sustain failures of half the nodes (rounding up) minus one if the disk witness goes offline or fails. For example, a six node cluster with a failed disk witness could sustain two (3-1=2) node failures.

    In my case, I have a 2 node 2008 cluster with Node and Disk Majority

    I took down the Quorum drive and the entire cluster went down

    Normal?

    No, as 2 out of 3 votes would have still been available.

    Having said that, in your config, removal of the disk witness would not sustain any further node failures.

    Re run the cluster validation report and check for errors, also check the event and cluster logs for issues too

    Thanks

    What I forgot to mention is that the Cluster went down and then came back up again after a few seconds. The Quorum drive also came up

    Thanks

  • October 31, 2012 at 4:52 am

    #1554537

    check your event logs for more info, there'll be some errors there that will be relative.

    -----------------------------------------------------------------------------------------------------------

    "Ya can't make an omelette without breaking just a few eggs" 😉

  • October 31, 2012 at 4:57 am

    #1554540

    Perry Whittle (10/31/2012)

    check your event logs for more info, there'll be some errors there that will be relative.

    Thanks for your help

  • October 31, 2012 at 8:37 am

    #1554639

    you're welcome 😉

    -----------------------------------------------------------------------------------------------------------

    "Ya can't make an omelette without breaking just a few eggs" 😉

Viewing 8 posts - 1 through 7 (of 7 total)

You must be logged in to reply to this topic. Login to reply