Several pitfalls during CEpH-deploy CEpH deployment in fedora 19

Recently, CEpH 0.81 was deployed on a virtual machine of Fedora 19.

Since CEpH has a simplified Deployment Tool CEpH-deploy, it seems that deployment is a very simple process. In theory, it should be the following steps:

0. In the experiment environment, delete all iptables rules, establish SSH trust between all machines, and add the CEpH Yum repo.

1. CEpH-deploy new mon osd1 osd2 osd3

2. CEpH-deploy install mon osd1 osd2 osd3

3. CEpH-deploy mon create-initial mon

4. CEpH-deploy gatherkeys mon osd1 osd2 osd3

5. CEpH-deploy OSD prepare osd1:/DIR osd2:/DIR osd3:/Dir

6. CEpH-deploy OSD activate osd1:/DIR osd2:/DIR osd3:/Dir

Should it be so simple? In actual operations, I found that the following needs to be added:

1. Several software packages required by CEpH need to be installed on fedora 19 (why can't CEpH-deploy check it by yourself ?) : RedHat-LSB, leveldb needs to be upgraded to 1.12, And the OSD node directly uses the block device to install hdparm.

2. Before step 5 of OSD prepare, You need to manually copy the bootstrap_osd from the Mon node to the OSD node. I forgot the complete path...

3. Before activating OSD, manually create an empty/var/lib/CEpH/OSD directory on the OSD node.

I wish you all a pleasant experience.

This article is from the computing-storage-Internet blog, please be sure to keep this source http://besmart.blog.51cto.com/8957700/1551760

Several pitfalls during CEpH-deploy CEpH deployment in fedora 19