xCAT/xcat-core
2
0
Fork 0
mirror of https://github.com/xcat2/xcat-core.git synced 2024-11-21 09:02:00 +00:00
Code Issues Releases Wiki Activity
Page: NFS_redundancy
Pages
2014_10_01_Bash_vulnerability 2014_11_04_Poodle_Attack AIX_NIM_NFSv4_support Add_Extra_Parameters_for_confignics_Postscript CRHS like_function CRHS like_function_enhancements CSM_CFM like_function_to_sync_config_files_to_the_nodes Capture_Linux_Image_from_running_node_to_create_a_stateless statelite_image Chef_databag_support_for_deploying_openstack_in_xCAT Cluster_config_file Conserver_Replacement Coordinated_Cluster_Shutdown_and_Bringup Copycds_enhancement Coral PDU supports Create_osimage_definition_for_Management_Node Cumulus_OS_Upgrade_Support_Test DFM_Service_Node_Hierarchy_support DHCPv6_management DNS_Enhancments DNS_master_slave_support Def_commands_to_display set_nic_attributes_more_easily Deploying_OpenStack_Baremetal_Nodes_Using_xCAT Disable_ssh_root_passwordless_node to node_access Document_large_cluster_tunables,_etc. Dynamic_node_groups_support Energy_Management_for_Flex Energy_management_support_for_blade Enhance_NTP_setup_support Enhancement_of_lsslp_ flexdiscover Enhancement_on_makeroutes_and_snmove_commands Exporting_more_table_attributes_to_nodes Filling_the_osimage_attributes_with_the_files_that_are_found_in_the_search_path Full_Automation_of_DB2_Setup_on_Service_Nodes Function Verification Test Design of xCAT High Availability GPFS_statelite_nodes Getmacs_and_rnetboot_enhancement Getnodetype_performance_imprvoments_ _phase_I HA_NFS_on_AIX_service_nodes HA_Service_Nodes_for_AIX HA_Service_Nodes_for_AIX_ _using_External_NFS_Server HFI_MAC_address_failover HPC_Integration_Phase_1 Hardware_Inventory_for_Node Health_Check_Script_Framework Hierarchical_Design Home Image_Provisioning Image_import_and_export_enhancement_for_Linux Improve_perl_array_search_efficiency_for_scalability Improvement_of_pre creating_mypostscripts_on_linux Initrd_Enhancement_separate_initrd_hacking_from_noteset_get_drivers_from_os_update_distro Kdump_for_Linux_diskless_nodes Linux_AIX_mixed_Clusters_postscripts_cleanup Local_Disk_support_for_statelite Lstree_for_Servicenode_Hardware_and_VM_hierarchies Management_Node_Pool Management_Node_Setup_Verification Managing_Multiple_Sites_and_Clusters Mellanox_Switch_Support Mini design for openbmc python framework Mini design_of_updating_pre defined_groups_for_nodes Minimize_the_install setup_steps_of_xcat_on_the_MN Mkvlan_enhancements Multiple_Zone_Support2 Multiple_pkgdir_support_with_full_installation NFS_Redundancy_Investigation NFS_redundancy New_Cloud_Tables_in_xCAT Node_Status_and_Application_Status Node_definition_installation_service_node_setup_verification Nodeset_defaults_when_node_provmethod=osimage_is_set OS_Deployment_Monitoring_And_Automatic_Retry OS_Distro_Update_xCAT_command_osdistroupdate OpenStack Chef Cookbook_xCAT_Integration OpenStack_Deployment_using_Chef OpenStack_Deployment_using_Puppet Ospkgs_enhancements_to_install upgrade_OS_packages_from_Updates Overview_of_HA_Services_for_Diskless_AIX_Nodes Power8_BE_support Predict_long_nicname_for_RH7 Proposal_for_docker_Support_in_xCAT RHEL7_Support RHEV_support Rcons_hierachy_support Redfish_support Refine_the_usage_of_installnic_and_primarynic_attributes Release processs automation investigation Resolving_Nodenames_for_System_p_HW_Ctrl Run_parallel_commands_for_Ethernet_switches Run_postbootscripts_after_each_diskful_reboot SLES12_Support SLES_11_SP1_SDK_support Secure root support Sequential_Discovery Sequential_Discovery_phase_2 Service_Node_Scenarios_and_Retry_Logic_in_Postscript_Process Service_node_manual_takeover_on_AIX Service_node_take_over Setup_IP_routes_between_the_MN_and_the_CNs Shared data based xCAT MN HA mini design Shared data based xCAT MN HA user case Shared data based xCAT MN HA Specify_the_disk_partitions_during_diskfull_install Steps to install xcat_openbmc_py3 on RH8 Support uploading openbmc firmware in parallel Support_Management_Node_in_the_servicenode_table Support_for_Redundant_BPAs_and_FSPs Support_multiple_hardware_control_points Support_osimage.pkgdir_with_multiple_paths_in_genimage_and_diskfull_installation Support_postscripts_associated_with_rsync_files_to_be_run_after_rsync_in_xdcp Support_to_inject_drivers_from_Driver_RPM_Package Support_xcatd_running_user_defined_sql_scripts_on_start_and_foreign_keys Supporting_driver_update_disk_for_Linux Supporting_the_scalable_functions_for_HX5_blade Switch_Discovery_And_Switch_Management System_p_Discovery_Enhancements System_z_hierarchical_tree Test Design of HA Test Design of TestCase Label Test Design of `goconserver` Regression Testing Test Design of `goconserver` Support Test Design of configib Test Design of cumulus osimage Test Design of cumulus_osimage Test Design of gonconserver redirect logs Test Design of hardware control commands regression Test Design of makegocons option l Test Design of rbeacon python version Test Design of rcons based goconserver Test Design of revenlog python version Test Design of rflash python version Test Design of rinv python version Test Design of rpower python version Test Design of rsetboot python version Test Design of rspconfig python version Test Design of rvitals python version Test Design of syncfiles Test Design of xcat inventory diff Test Design of xcat inventory export import performance Test Design of xcat inventory option e Test Design of xcat inventory osimage environment variables export part Test Design of xcat inventory osimage environment variables Test Design of xcat inventory osimage phase2 Test Design of xcat inventory osimage support share file directory Test Design of xcat inventory osimage Test Design of xcat inventory phase 1 Test Design of xcat inventory regular expression Test Design of xcat inventory save keys Test Design of xcat inventory user case Test Design of xcat terraform provider phase1 Test Report of `goconserver` Test cases against nodeset error handling Test_Case_Designs_FVT The mini design of osimage version control(initial draft) The mini design of syncfiles for cumulus The mini design of using xcat inventory templates to setup xCAT cluster with ansible The_remote_console_support_of_hmc Trace_Facility_for_xCAT Ubuntu_supports_internet_repository Updatenode_Performance_Enhancements Updatenode_Sync_Service_Nodes_for_install Updatenode_enhancement_for_security_keys_and_certificates Updatenode_support_to_run_mypostscript_sections Updatenode_to_the_Management_Node Updates_to_xcatsetup_and_Cluster_Config_File Windows_Multiple_Disks_Partitions_Configuration Windows_Multiple_Nics_Configuration Windows_Multiple_Winpes Windows_Support_to_Run_Postscripts_Postbootscripts_in_Windows_Deployment Wish_List_for_xCAT_2 Wish_List_for_xCAT_FVT XCAT 2 Mini Designs for New Features XCAT_2.10_Release_Notes XCAT_2.11.1_Release_Notes XCAT_2.11_ESP_Release_Notes XCAT_2.11_Release_Notes XCAT_2.12.1_Release_Notes XCAT_2.12.2_Release_Notes XCAT_2.12.3_Release_Notes XCAT_2.12.4_Release_Notes XCAT_2.12.5_Release_Notes XCAT_2.12_Release_Notes XCAT_2.13.10_Release_Notes XCAT_2.13.11_Release_Notes XCAT_2.13.1_Release_Notes XCAT_2.13.2_Release_Notes XCAT_2.13.3_Release_Notes XCAT_2.13.4_Release_Notes XCAT_2.13.5_Release_Notes XCAT_2.13.6_Release_Notes XCAT_2.13.7_Release_Notes XCAT_2.13.8_Release_Notes XCAT_2.13.9_Release_Notes XCAT_2.13_Release_Notes XCAT_2.14.1_Release_Notes XCAT_2.14.2_Release_Notes XCAT_2.14.3_Release_Notes XCAT_2.14.4_Release_Notes XCAT_2.14.5_Release_Notes XCAT_2.14.6_Release_Notes XCAT_2.14_Release_Notes XCAT_2.15.1_Release_Notes XCAT_2.15_Release_Notes XCAT_2.16.1_Release_Notes XCAT_2.16.2_Release_Notes XCAT_2.16.3_Release_Notes XCAT_2.16.4_Release_Notes XCAT_2.16.5_Release_Notes XCAT_2.16_Release_Notes XCAT_2.17_Release_Notes XCAT_2.9.2_Release_Notes XCAT_2.9.2_aix_Release_Notes XCAT_2.9.3_Release_Notes XCAT_2.9.4_Release_Notes XCAT_Cluster_in_SoftLayer XCAT_HPC_Integration_Files_per_OS,_Architecture,_Product XCAT_Health_Check XCAT_IPv6_support_on_AIX XCAT_IPv6_support_on_Linux XCAT_Mini Design_Template XCAT_Monitoring_Failure_Analysis_Conditions_Batched_Events XCAT_Postscripts_framework_enhancements XCAT_Provisioning_OS_on_consistent_disk XCAT_REST_API_enhancements XCAT_Support_for_Kits_and_Kit_Components XCAT_probe XCAT_support_for_multiple_network_domains Xcatexport_and_xcatimport_commands Xcatmon_and_appstatus_for_HPC_products Xeon_Phi_support cofignetwork_for_RH7 confignetwork integrate confignics_configeth_configib functions usage determine_install_disk diskdiscover_and_configraid docker_instances_lifecycle_management enhancement_for_rinstall_command logging pre script to make debug easy makedhcp_append_statement_automatically_rather_than_get_statement_from_nodeset mini design of ZTP based onie switch discovery and configuration minidesign Switch based switch discovery minidesign_xcatprobe node status enhancement ntp_enhancement openbmc_support post.xcat_restructure rflash_streamline_procedure rscan command for hypervisor rspconfig network rspconfig ntpservers rspconfig password scaleup_basedon_xcat_in_openshift_cluster supporting_partitionfile_for_Ubuntu supporting_persistent_kernel_cmdline_after_diskful_installation test_sidebar the mini design of xcat inventory export and import for osimage use fifo pipe to enhance getadapter xCAT Build Schedule xCAT_2.12_ReleaseInformation xCAT_2.13_ReleaseInformation xCAT_2.14_ReleaseInformation xCAT_2.15_ReleaseInformation xCAT_Docker_Image xcatprobe_framework xcatprobe_requests
8 NFS_redundancy
ligc edited this page 2015-07-30 04:22:49 -04:00
Table of Contents

Table of Contents

{{:Design Warning}}

Mini-design for External NFS Support with AIX stateless/statelite and Linux statelite

xCAT AIX stateless/statelite and Linux statelite compute nodes need to mount NFS directories from the service node, each service node can serve hundreds of compute nodes, if the NFS service on the service node or the service node itself runs into problem, all the compute nodes served by this service node will be taken down immediately, so we have to consider providing the redundant NFS service for the compute nodes.

The reason we are focusing on NFS and not all of these other services on the service nodes, is because NFS is the only one that will take the nodes down immediately and force a reboot. For most of the other services, the administrator will have a little while to run the utility to switch the nodes to another service node. For dhcp, the lease time can be made longer. For dns, they can replicate /etc/hosts if they want.

There are a lot of ways to achieve the HA NFS for the stateless/statelite compute nodes, from xCAT perspective, we will focus on two methods:

  1. External NFS server: the external NFS server is setup with HA capability, for example, SONAS or GPFS CNFS, and the required NFS exports for the stateless/statelite image are also configured on the external NFS server, the stateless/statelite compute nodes will mount the NFS exports from the external NFS server instead of their management node or service nodes, so if any of the service node fails, it will not take the nodes down immediately.

  2. Setup HA NFS server on service nodes: Use any existing HA product such as IBM Tivoli System Automation to setup HA NFS server between service nodes, the HA NFS service will be provided through a service IP address, the service IP address will be failovered if the current primary server goes down, the stateless/statelite compute nodes will mount the NFS exports through the service IP address.

This documentation describes external NFS support with AIX stateless/statelite and Linux statelite. The configuration of HA NFS on service nodes will be addressed in a separate documentation.

1. Architectural Diagram

The core idea of this external NFS server support is to redirect all the NFS mounts on the stateless/statelite nodes to an external NFS server, the external NFS server can be any configuration as long as it could provide HA NFS capability and the performance is not a problem when thousands of NFS mounts come from all the stateless/statelite nodes.

The diagram below illustrates the architectural structure of the external NFS server support, all the compute nodes will mount NFS directories only from the external HA NFS server, the service nodes will continue to provide the other network services such as TFTP, DHCP to the compute nodes.

When any of the service node fails, the compute nodes served by it will not be affected immediately, the applications can continue run on the compute nodes. But the administrator still needs to recover the failed service nodes very soon, because the service nodes are still providing network services to the compute nodes, for example, the DHCP lease expire before the service node is recovered will cause problem.

2. External Interface

The existing node attribute “nfsserver” is used to specify the external nfs server, the procedure needed for external NFS server support includes:

1. Set the node attribute “nfsserver” to hostname of the external NFS server before running mkdsklsnode/nodeset.

2. After the mkdsklsnode/nodeset command is run:

   i. (Optional) Verify the operating system image and the other settings by booting up one of the nodes.


   ii. Copy the required operating system image files from the service node(hierarchy cluster) or management node(non-hierarchy)to the external NFS server.

When copying the operating system image files, the files modes should be preserved, thus you should specify the “preserve” option with the remote copy or remotesynchronization commands, take rsync as an example, the “-az” should be the correct option. Of course, any backup/restore commands such as tar or zip can be used to copy the operating system image files.

To avoid confusion for both operating system deployment tools(NIM, kickstart and autoyast) and the users, it is required that the operating system image directories structure should be the same on the service nodes and the external NFS server, for example, the SPOT of an AIX image is under directory /install/nim/spot/aix71/usr on the service nodes, then the SPOT should also be copied to the directory /install/nim/spot/aix71/usr on the external NFS server.

   iii. Setup the NFS exports on the external NFS server.

Please be aware that the NFS export options for the AIX image directories on the external NFS server should be the same with the ones on the service nodes, a simple way is to copy the /etc/exports from the service nodes to the external NFS server.

If the diskless nodes are running AIX and the external NFS server runs Linux, when setting up NFS exports on the external NFS server, the “insecure” exports option is needed to make the AIX be able to mount the Linux NFS exports.

   iv. If any changes are made on the management node or service nodes and the mkdsklsnode/nodeset command needs to be run again, the operating system image files need to synchronized to the external NFS server.

The operating system image files are different for AIX non-shared_root and shared_root configuration; there are some additional steps with the statelite support, the following sections will cover the procedure for AIX non-shared_root stateless, AIX shared_root statelss, AIX statelite and Linux statelite.

2.1 AIX non-shared_root stateless

The following operating system image files need to be copied to the external NFS server for the AIX non-shared_root stateless configuration:

1. SPOT

Use lsnim -l <spot_name> to get the location of the SPOT, for example, /install/nim/spot/71Bcosi/usr, copy the whole directory to the external NFS server.

For one AIX stateless image, the SPOT resource is the same across all the service nodes, so only needs to copy the SPOT from one service node to the external NFS server. Here is an example:

On one service node:

#rsync -az /install/nim/spot/71Bcosi/usr <external_nfs_server>:/install/nim/spot/71Bcosi/

2. ROOT

Use lsnim -l <root_name> to get the location of the root resource, for example, /install/nim/root/71Bcosi_root.

For the non-shared_root stateless configuration, each node has a separate root resource, all the root resources are under the same directory, you can copy the directory that contains all the root resources to the external NFS server.

Even for the same AIX image, the root resource is different between service nodes, so the root resource directories from all the service nodes are required to be copied to the external NFS server. Here is an example:

On all the service nodes:

#rsync -az /install/nim/root/71Bcosi_root <external_nfs_server>:/install/nim/root/

3. Setup NFS on the external NFS server

Since all the root directories for all the nodes should be exported, so there will be a lot of NFS export entries on the external NFS server, a simple way is to put the content of all the /etc/exports files on all the service nodes into the /etc/exports file on the external NFS server and then run exports -a to export all the required directories.

2.2 AIX shared_root stateless

1. SPOT

Use lsnim -l <spot_name> to get the location of the SPOT, for example, /install/nim/spot/71Bcosi/usr, copy the whole directory to the external NFS server. For one AIX stateless image, the SPOT resource is the same across all the service nodes, so only needs to copy the SPOT from one service node to the external NFS server. Here is an example:

On one service node:

#rsync -az /install/nim/spot/71Bsharedrootcosi/usr <external_nfs_server>:/install/nim/spot/71Bsharedrootcosi/

2. shared_root

Use lsnim -l <shared_root_name> to get the location of the root resource, for example, /install/nim/root/71Bcosi_root. For the shared_root stateless configuration, all the nodes share the same shared_root resource, but the shared_root resource will be updated for each node when running mkdsklsnode, all the updates for specific nodes are under the etc/.client_data subdirectory in the shared_root directory, so we can copy the shared_root resource from one service node to the external NFS server, and then copy all the etc/.client_data subdirectories from all the service nodes to the external NFS server. Here is an example:

Here is an example:

On one service node:

#rsync -az /install/nim/shared_root/71Bsharedrootcosi_shared_root <external_nfs_server>:/install/nim/shared_root/

On all the service nodes:

# rsync -az /install/nim/shared_root/71Bsharedrootcosi_shared_root/etc/.client_data <external_nfs_server>:/install/nim/shared_root/71Bsharedrootcosi_shared_root/etc/

3. Setup NFS on the external NFS server

All the nodes will use the same shared_root directory, so only one NFS exports entry is required for the shared_root configuration, however, the NFS exports options are different on each service node, for example, the “root=” and the “access=” are used to specify the nodes list served by each servicde node, so you still need to put the content of all the /etc/exports files on all the service nodes into the /etc/exports file on the external NFS server and then run exports -a to export all the required directories.

<Designer Note:> 1. Should the paging resource be served from the external NFS server? _ The paging space will not be used in most of the time, xCAT can configure the paging resource as sparse_paging, we also did some experiment and it turned out that the AIX diskless client can keep running normally when we remove the paging file manually after the diskless client was up and running. So the current plan is not to failover the paging resource to avoid too many NFS exports on the external NFS server._

2. For now we only put the root/shared_root and the SPOT on the external NFS server, should we consider putting the other NIM resources such as dump and user defined resources to the external NFS server?

3. Is it necessary to provide external interface to users to specify which NIM resources should be put on the external NFS server?''

2.3 AIX statelite

AIX statelite is actually the shared_root stateless plus “overlay” for specific files or directories through the information in tables statelite, litefile and litetree. So the external NFS support for AIX statelite is similar with the shared_root stateless configuration, except that the considerations for the information in tables statelite, litefile and litetree.

statelite table

The “statemnt” column in the statelite table specifies the persistant read/write area where a node's persistent files will be written to, the persistent files are usually critical for applications running on the nodes, so it is recommended to point the “statemnt” to the external NFS server, the variable $noderes.nfsserver can be used in the statelite table to specify that the node's attribute “nfsserver” is the external NFS server. Here is an example of the statelite table:

node,image,statemnt,comments,disable

"aixcn1",,"$noderes.nfsserver:/stateliteroot",,__

Make sure you NFS export any directories that are listed in this table with read-write permission on the external NFS server before attempting to boot the nodes.

litefile table

The litefile table specifies the directories and files for the statelite setup along with the option to use to do the setup. All the directories specified in the liefile table should also be created and set to correct permission on the external NFS server before attempting to boot the nodes.

litetree table

The litetree table controls where the initial content of the files in the litefile table come from, and the long term content of the “ro” files. If any directories in this table point to the external NFS server, make sure you NFS export them on the external NFS server before attempting to boot the nodes.

2.4 Linux statelite

Linux statelite nodes need to mount the rootimg from the service nodes, to provide high availability NFS capacity, the external NFS server should serve the rootimg for the Linux statelite nodes.

1. Before running nodeset osimage=, change the nodes attribute nfsserver to the hostname or ip address of the external NFS server.

2. After nodeset osimage= is run, sync the rootimg directory to the external NFS server, here is an example:

rsync -az /install/netboot/rhels5.5/ppc64/compute 9.114.47.101:/install/netboot/rhels5.5/ppc64/

The syntax and definition of the tables statelite, litefile and litetree is same for Linux statelite and AIX statelite, so the considerations for these three tables are the same for Linux and AIX, please refer to the "AIX Statelite" section above for the detailed procedure.

News

History

  • Oct 22, 2010: xCAT 2.5 released.
  • Apr 30, 2010: xCAT 2.4 is released.
  • Oct 31, 2009: xCAT 2.3 released.
    xCAT's 10 year anniversary!
  • Apr 16, 2009: xCAT 2.2 released.
  • Oct 31, 2008: xCAT 2.1 released.
  • Sep 12, 2008: Support for xCAT 2
    can now be purchased!
  • June 9, 2008: xCAT breaths life into
    (at the time) the fastest
    supercomputer on the planet
  • May 30, 2008: xCAT 2.0 for Linux
    officially released!
  • Oct 31, 2007: IBM open sources
    xCAT 2.0 to allow collaboration
    among all of the xCAT users.
  • Oct 31, 1999: xCAT 1.0 is born!
    xCAT started out as a project in
    IBM developed by Egan Ford. It
    was quickly adopted by customers
    and IBM manufacturing sites to
    rapidly deploy clusters.