Forums

Home / Forums

You need to log in to create posts and topics. Login · Register

error deploy second node

when i try to deploy lsecond node i receive error on:

Error binding cluster

Create ceph monitor on 192.168.101.200 failed.

what i wrong ? i stay in a virtual lab with vmware

i have 4 cpu and 10 gb of ram for each node and 5 nic

The cluster should be built after the deployment of 3rd node not the second.  Maybe there was a earlier attempt , i would recommend re-installing and see if this fixes it. If after installing you still get same issue please send me the log files for all 3 nodes found in  /opt/petasan/log/PetaSAN.log to admin @ petasan.org

If you run under vmware ESXi, after installing go in the node console (blue screen) menu -> interface naming to rename/re-oder you nics to match the order in ESXi. ESXi has a strange issue when adding the 5th nic it inserts it as eth1 instead of eth4

i retry with 3 nic and work fine. thanks so much.

 

just one question.. why i m able to create 4tb space if i have much less ?

in add.. can i add hard disk to my single nod to increase avaible storage ?

 

can i see on whic disc there are my data ?

just one question.. why i m able to create 4tb space if i have much less ?

You can over commit / thin provision storage. Your clients can see more disk space in their iSCSI disks than  physical disk capacity . The important thing is as these clients start writing and filling the disks then you will need to add physical disks on a per needed basis. There are charts on and warnings in the dashboard when the cluster starts to fill, you will also get email alarms if you configured email notifications.

in add.. can i add hard disk to my single nod to increase avaible storage ?

Yes you can add physical disks to increase your physical storage. Adding physical disks also increases performance of your iSCSI disks.

can i see on whic disc there are my data ?

Each iSCSI disk uses all physical disks, so its data is distributed across all physical disks, if you add a new physical disk it will automatically get pieces of each existing disk so that the data of the iSCSI disk is distributed evenly. Adding physical disks will also make your exiting iSCSI faster. To get into more detail each 4M sector/stripe of your iSCSI disk is stored in 2 or 3 ( depending on your replica count) different physical disks.

thanks so much.

 

Then sorry for my question but i evaluated product...

 

now..

 

can i remove a node ?

 

You can remove a non management node ( ie node 4 onward) if it is not running (fails/shut down), you do so from the Node List page.

You can not remove  management node ( node 1 to 3 ) . If a management node fails, you need to replace it soon, the cluster needs at least 2 management nodes to fucntion. Replace management node is the 3rd option on the first page of the Deployment Wizard.