Nagios Support Forum

Just wanted to bring this to your attention.

I went through the documentation for resizing the HDD's for our 3 host NLS cluster. Of the three hosts, only one had no issues when resizing the disk. The other two would not boot after the cfdisk operations were performed and posted an "ERROR 17" on boot.

https://assets.nagios.com/downloads/nag ... M-Disk.pdf

Thats the documentation I was following (CentOS6 appliance). Steps taken were to shut down the Linux VM, expand the disk in vmware, boot the VM back up, run cfdisk and its operations, then reboot. On reboot after cfdisk is when we ran into the issues, this is before resize2fs. I only did a single host at a time, giving the cluster time to rebuild after each operation.

I had to boot into a rescue disk and repair the partition with testdisk, then I just expanded them using gparted and everything has been fine. Not sure whats changed but wanted to give feedback on what I ran into when using the official documentation.

Hello Ehamby ,

Given the fact that you were able to resize the disk successfully on one of your hosts in addition to failing on the other two, it's difficult to know how to proceed with this. The steps in the documentation have been thoroughly tested and we will continue to do so. Please describe the differences, if any, between the partition tables on these hosts. If you've resolved this on your own, may we lock the thread?

bolson wrote:Hello Ehamby ,

Given the fact that you were able to resize the disk successfully on one of your hosts in addition to failing on the other two, it's difficult to know how to proceed with this. The steps in the documentation have been thoroughly tested and we will continue to do so. Please describe the differences, if any, between the partition tables on these hosts. If you've resolved this on your own, may we lock the thread?

All three of the hosts were deployed using the official VM OVF appliance. The cluster was originally a two host cluster, which we eventually added a third host to. The one that worked, was the third host. The two that failed, were the original hosts in the cluster. All hosts had 350GB of HDD, which was expanded to 450GB. Other than that, there's nothing special about the VM's or the partition tables.

I'm sure its been tested thoroughly, I just wanted to relay the issues I ran into when expanding the disks with cfdisk. Feel free to close it out if you want.

Thanks for the feedback. I have a few quick questions before locking this up: what was the initial version of NLS installed on the two machines that failed? Where they upgraded? Was the third appliance from the same ovf ?

cdienger wrote:Thanks for the feedback. I have a few quick questions before locking this up: what was the initial version of NLS installed on the two machines that failed? Where they upgraded? Was the third appliance from the same ovf ?

Sorry for the delay, I don't recall the exact release number but the first two VM's would have been one of the first releases. The the third VM was added last year (two years after the first two VM's were created on our side). When I added the third VM I redownloaded the appliance from the NLS website directly, and upgraded the existing VM's to the version of the new VM prior to adding it to the cluster.

Thanks! I suspect it has something to due to the earlier versions. Glad you were able to find a work around! Locking now.