After getting NFSv4 working, it’d be of course nice to automatically mount the nfs exported home directories. In this part I’m going through the steps to get school specific home directories mounted from a central server. Using autofs is an alternative to defining the mounted directories in /etc/fstab. It mounts the directories automatically when they are needed instead of doing it at boot time.After getting NFSv4 working, it’d be of course nice to automatically mount the nfs exported home directories. In this part I’m going through the steps to get school specific home directories mounted from a central server. Using autofs is an alternative to defining the mounted directories in /etc/fstab. It mounts the directories automatically when they are needed instead of doing it at boot time.After getting NFSv4 working, it’d be of course nice to automatically mount the nfs exported home directories. In this part I’m going through the steps to get school specific home directories mounted from a central server. Using autofs is an alternative to defining the mounted directories in /etc/fstab. It mounts the directories automatically when they are needed instead of doing it at boot time.
Next it’s time to finally get files moving between the servers. For this we use NFSv4 that supports kerberos out-of-the-box also in Ubuntu. This part is based on the newest Lucid packages in the repositories, which should be pretty close to alpha 3 now.Next it’s time to finally get files moving between the servers. For this we use NFSv4 that supports kerberos out-of-the-box also in Ubuntu. This part is based on the newest Lucid packages in the repositories, which should be pretty close to alpha 3 now.Next it’s time to finally get files moving between the servers. For this we use NFSv4 that supports kerberos out-of-the-box also in Ubuntu. This part is based on the newest Lucid packages in the repositories, which should be pretty close to alpha 3 now.
Kerberos requires every client to know where the server is located. This can be done either by using /etc/krb5.conf file or using DNS to distribute the information. Using DNS makes it easier to do changes in the network settings as not every client needs to be updated. Next we aim to minimize the amount of configuration needed for every client so configuring DNS properly is a logical first step.Kerberos requires every client to know where the server is located. This can be done either by using /etc/krb5.conf file or using DNS to distribute the information. Using DNS makes it easier to do changes in the network settings as not every client needs to be updated. Next we aim to minimize the amount of configuration needed for every client so configuring DNS properly is a logical first step.Kerberos requires every client to know where the server is located. This can be done either by using /etc/krb5.conf file or using DNS to distribute the information. Using DNS makes it easier to do changes in the network settings as not every client needs to be updated. Next we aim to minimize the amount of configuration needed for every client so configuring DNS properly is a logical first step.
After getting OpenLDAP running properly and the schemas in place, the next step is to get Kerberos and AutoFS running on top of it to enable centrally managed automatic NFSv4+kerberos mounts to user home directories. Here we setup kerberos using OpenLDAP as the backend to store the principals. This allows one to easily replicate the data to slave servers.After getting OpenLDAP running properly and the schemas in place, the next step is to get Kerberos and AutoFS running on top of it to enable centrally managed automatic NFSv4+kerberos mounts to user home directories. Here we setup kerberos using OpenLDAP as the backend to store the principals. This allows one to easily replicate the data to slave servers.After getting OpenLDAP running properly and the schemas in place, the next step is to get Kerberos and AutoFS running on top of it to enable centrally managed automatic NFSv4+kerberos mounts to user home directories. Here we setup kerberos using OpenLDAP as the backend to store the principals. This allows one to easily replicate the data to slave servers.
In this part I’m setting up ldap schemas for samba, autofs and kerberos. This is needed before the actual configuration for these can be done. Unfortunately I could not find ldif files for OpenLDAP for these, so the schema files need to be converted to ldif files. The tutorial at help.ubuntu.com instructs to use the slaptest tool for this.In this part I’m setting up ldap schemas for samba, autofs and kerberos. This is needed before the actual configuration for these can be done. Unfortunately I could not find ldif files for OpenLDAP for these, so the schema files need to be converted to ldif files. The tutorial at help.ubuntu.com instructs to use the slaptest tool for this.In this part I’m setting up ldap schemas for samba, autofs and kerberos. This is needed before the actual configuration for these can be done. Unfortunately I could not find ldif files for OpenLDAP for these, so the schema files need to be converted to ldif files. The tutorial at help.ubuntu.com instructs to use the slaptest tool for this.
