11[[multicluster]]
2- == Multi -node cluster setups for {sapS4}
2+ == Setting up multi -node cluster setups for {sapS4}
33
4- *Multi-node cluster* setups mean cluster configurations with more than two nodes.
5- Depending on the starting point it is possible to extend a two-node cluster setup
4+ *Multi-node cluster* setups are cluster configurations with more than two nodes.
5+ Depending on the starting point, you can or extend a two-node cluster setup
66or directly start with more than two nodes for an ASCS / ERS high availability setup.
7- The examples below will show the setting up of multi-node cluster and the extension
8- of an existing two node cluster pair. The major configuration changes will be shown
9- and the basic preparation of the new cluster member node.
10-
11- The task list to set up the three node cluster is similar to the task list for the
12- two-node cluster. However, some details are described different here to get a diskless
13- SBD setup. Such a diskless SBD setup is an optional improvement for three nodes, but
14- does not work for two nodes. On the other hand, priority fencing is an optional improvement
15- for two nodes, but does not work for three nodes. An example priority fencing configuration
16- for the two-node cluster is shown in the appendix.
7+ The following examples show how to set up a multi-node cluster and how to extend
8+ an existing two-node cluster. They outline major configuration changes and the basic
9+ preparation of a new member node.
10+ The task list for setting up a three-node cluster is similar to the task list for a
11+ two-node cluster. However, some details differ to allow for a diskless
12+ SBD setup. A diskless SBD setup is an optional improvement for three nodes, but
13+ does not work on two nodes. Conversely, priority fencing is an optional improvement
14+ for two nodes, but does not work on three nodes. An example priority fencing configuration
15+ for a two-node cluster is available in the appendix.
1716See the {sleHA} product documentation for details.
1817// TODO PRIO2: fix product links
1918// (https://documentation.suse.com/sle-ha/15-SP3/single-html/SLE-HA-administration/#pro-ha-storage-protect-fencing).
@@ -23,18 +22,18 @@ NOTE: When extending a cluster from two to three nodes, make sure to not use pri
2322////
2423.Tasks
2524
26- . OS preparation and install the cluster software
25+ . Preparing the OS and installing the cluster software
2726. Preparing watchdog and SBD parameters for diskless SBD
28- . Configure the cluster base including corosync and resource manager with diskless SBD
29- . Configure the cluster resources
30- . Tune the cluster timing in special for the SBD.
27+ . Configuring the cluster base including corosync and resource manager with diskless SBD
28+ . Configuring the cluster resources
29+ . Tuning the cluster timing in special for the SBD.
3130////
3231
3332////
3433// BEGIN OF BLOCKED CHAPTERS FOR MULTINODE 2019-06-18
3534=== Setting up multi-node cluster from the top with diskless SBD
3635
37- A fencing mechanism is essential for cluster controlled infrastructures. Sometimes there is no possibility
36+ A fencing mechanism is essential for cluster- controlled infrastructures. Sometimes there is no possibility
3837to have a shared block device or an IPMI device. A solution for such environments could be the
3938diskless SBD option. SBD can be operated in a diskless mode. In this mode, a watchdog device will be
4039used to reset the node in the following cases: if it loses quorum, if any monitored daemon is lost and not
@@ -48,7 +47,7 @@ SBD in diskless mode cannot handle split brain scenarios for two-node clusters.
4847==== Preparing the operating system and installing the cluster software
4948
5049In general the OS preparation is similar to the procedure described for the two node cluster.
51- This includes the setup of NTP as well as installing the software pattern * ha_sles* using * yast2* or * zypper* .
50+ This includes the setup of NTP as well as installing the software pattern ` ha_sles` using ` yast2` or ` zypper` .
5251
5352==== Preparing watchdog and SBD parameters for diskless SBD
5453
@@ -219,14 +218,14 @@ for the two node cluster.
219218. Patching the existing nodes
220219. Preparing the new node's operating system
221220. Installing the cluster software on the new node
222- . Preparing SAPStartSrv resource agent on the new node
221+ . Preparing * SAPStartSrv* resource agent on the new node
223222. Preparing the SAP installation on the new node
224223. Adding the new node to the cluster
225224. Testing the new cluster configuration
226225
227226==== Backing up the current cluster
228227
229- To *back up the current cluster*, perform a
228+ To *back up the current cluster*, perform a
230229==========
231230
232231* Backup of your system including
@@ -238,17 +237,17 @@ The system is configured as described in the SUSE Best Practices document
238237__SAP S/4 HANA - Enqueue Replication 2 High Availability Cluster With Simple Mount - Setup Guide__.
239238
240239To *back up the cluster configuration*, go to one of the cluster nodes and save
241- the cluster configuration with _crm_ and _crm report_ commands:
240+ the cluster configuration with the _crm_ and _crm report_ commands:
242241
243242[subs="attributes"]
244243----
245244# crm configure save 2nodes_crm.txt
246245# supportconfig
247246----
248247
249- Back up the existing /etc/corosync/corosync.conf and all other files which may
250- be important for a restore. This example is one method for creating a backup.
251- The important point is using an external destination.
248+ Back up the existing _ /etc/corosync/corosync.conf_ and all other files required to restore the system.
249+ The following example is one method for creating a backup.
250+ The important step is storing the backup on an external destination.
252251
253252[subs="specialchars,attributes,quotes"]
254253----
@@ -265,25 +264,25 @@ The important point is using an external destination.
265264
266265==== Installing the operating system of the new node
267266
268- We recommend using automating the installation to ensure that the system setup
269- across nodes is identical. Make sure to document any additional steps you take
270- beyond the automated setup. In our example, we deploy the machines with Agama
271- and use Ansible for the configuration.
267+ We recommend using automated installation to ensure that the system setup
268+ across nodes is identical. Document any additional steps taken
269+ beyond the automated setup. In this example, we deploy machines with Agama
270+ and use Ansible for configuration.
272271
273272==== Patching the existing nodes
274273
275274If applicable, install the latest updates and patches on your existing nodes.
276- Alternatively, if you are using frozen repositories such as those provided by
277- Multi Linux Manager, add the new system to the same repositories, so they have
275+ Alternatively, if you are using frozen repositories, such as those provided by
276+ SUSE Multi Linux Manager, add the new system to the same repositories to ensure they have
278277the same patch level as your existing nodes.
279278
280- Use `zypper patch` or `zypper update` depending on your company's rules.
279+ Use `zypper patch` or `zypper update`, depending on your company's rules.
281280
282281- We recommend installing the latest available patches to guarantee system stability
283282and hardening. Bug fixing and security patches help avoid unplanned outages and make
284283the system less vulnerable.
285284
286- There are multiple ways:
285+ There are multiple ways to do so :
287286
288287[subs="specialchars,attributes,quotes"]
289288----
@@ -303,7 +302,7 @@ There are multiple ways:
303302# ping <hostname>
304303----
305304
306- - Set up * chrony* (this is best done with * yast2* ) and enable it:
305+ - Set up ` chrony` (this is best done with ` yast2` ) and enable it:
307306+
308307[subs="specialchars,attributes,quotes"]
309308----
@@ -337,19 +336,19 @@ same patch level and that all required reboots have been performed.
337336
338337============
339338
340- ==== Preparing SAPStartSrv resource agent on the new node
339+ ==== Preparing * SAPStartSrv* resource agent on the new node
341340
342- - Enable the sapping and sappong services.
341+ - Enable the ` sapping` and ` sappong` services.
343342+
344343[subs="specialchars,attributes,quotes"]
345344----
346345# systemctl enable sapping
347346# systemctl enable sappong
348347----
349348
350- After installing all necessary packages, compare installed package versions.
349+ After installing all necessary packages, compare the installed package versions.
351350
352- .Check that all nodes have the same software packages and versions:
351+ .Check if all nodes have the same software packages and versions:
353352============
354353- On the existing cluster nodes, type:
355354+
@@ -373,12 +372,12 @@ After installing all necessary packages, compare installed package versions.
373372----
374373============
375374
376- NOTE: If there any differences fix them first before you proceed.
375+ NOTE: If there are any differences, fix them first before you proceed.
377376
378377.Install and configure the watchdog device on the new machine.
379378============
380379
381- Instead of deploying the software-based solution, preferably use a hardware-based
380+ Instead of deploying the software-based solution, use a hardware-based
382381watchdog device. The following example uses the software device but can be easily
383382adapted to the hardware device.
384383
@@ -392,7 +391,7 @@ adapted to the hardware device.
392391============
393392
394393NOTE: Ensure that the new node is connected to the same SBD disk connected to
395- the existing two nodes. Ensure that the new node has the exact same SBD
394+ the existing two nodes. Ensure that the new node has identic SBD
396395configurations as the two existing nodes
397396
398397[subs="attributes"]
@@ -402,12 +401,12 @@ configurations as the two existing nodes
402401
403402==== Preparing the SAP installation on the new node
404403
405- With SWPM 2.0 (SP4 or later), which is part of the SL Toolset, SAP provides an new
404+ With SWPM 2.0 (SP4 or later), which is part of the SL Toolset, SAP provides a new
406405option which can perform all necessary steps to prepare a fresh install server to
407- be able to fit into an existing SAP system. This new option will help us to prepare
408- a new host which can later run either the ASCS or ERS in the cluster environment.
406+ be able to fit into an existing SAP system. This new option helps to prepare
407+ a new host which can later run either ASCS or ERS in the cluster environment.
409408
410- You need to create the directory structure that should run the SAP resource.
409+ Create the directory structure that should run the SAP resource.
411410The instance directory is located on an NFS share for all nodes.
412411
413412- Create mount points and mount NFS shares on the new added node ({my3nd3}):
@@ -464,10 +463,11 @@ The next step requires the following information:
464463# /usr/sap/hostctrl/exe/sapstartsrv pf=/usr/sap/<SID>/SYS/profile/<SID>_ASCS<instanceNumberAscs>_<virtHostNameAscs> -reg
465464----
466465+
467- NOTE: This must be done for each instance. Call sapstartsrv with parameters *pf=*<profile-of-the-sap-instance> and *-reg*.
466+ NOTE: This must be done for each instance.
467+ Call `sapstartsrv` with parameters *pf=*<profile-of-the-sap-instance> and *-reg*.
468468+
469469
470- - Disable * systemd* services of the ASCS and the ERS {sap} instance:
470+ - Disable ` systemd` services of the ASCS and the ERS {sap} instance:
471471+
472472[subs="specialchars,attributes"]
473473----
@@ -478,7 +478,7 @@ NOTE: This must be done for each instance. Call sapstartsrv with parameters *pf=
478478NOTE: This is mandatory for giving control over the instance to the HA cluster.
479479+
480480
481- - Check the {sap} * systemd* integration:
481+ - Check the {sap} ` systemd` integration:
482482+
483483[subs="specialchars,attributes"]
484484----
@@ -501,8 +501,8 @@ systemctl --no-ask-password start SAP{mySid}_{myErsIno} # sapstartsrv pf=/usr/sa
501501.Adding the New Node to the Cluster
502502==========
503503Check if the SBD device is available in case the SBD fencing method is in place
504- for the two nodes. If the existing cluster using a different, supported fencing
505- mechanism check and verify them too for the new cluster node.
504+ for the two nodes. If the existing cluster uses a different, supported fencing
505+ mechanism, check and verify it also for the new cluster node.
506506
507507[subs="specialchars,attributes,quotes"]
508508----
@@ -527,14 +527,14 @@ the file _/etc/corosync/corosync.conf_.
527527# corosync-quorumtool -s
528528----
529529
530- The values *expected_votes* and *two_node* should now look like this on all nodes:
530+ The values *expected_votes* and *two_node* should now look like the following on all nodes:
531531
532532----
533533expected_votes: 3
534534two_node: 0
535535----
536536
537- Modify the cluster configuration and set a new colocation rule with _crm_ :
537+ Modify the cluster configuration and set a new colocation rule with `crm` :
538538
539539[subs="specialchars,attributes,quotes"]
540540----
@@ -545,21 +545,21 @@ Modify the cluster configuration and set a new colocation rule with _crm_:
545545
546546==== Testing the new cluster configuration
547547
548- It is highly recommended to run certain test to verify that the new configuration
549- is working as expected. A list of test can be found in the basic setup for the two
548+ We recommend running test to verify that the new configuration
549+ is working as expected. Find a list of tests in the basic setup for the two
550550node cluster above.
551551
552552=== Pros and Cons for odd and even numbers of cluster nodes
553553
554554There are certain use cases and infrastructure requirements which end up in different
555- setups. We will cover some advantages and disadvantages of special configuration below :
555+ setups. Below, find a list of advantages and disadvantages of special configurations :
556556
557557* The two node cluster and two locations
558558 - Advantage: symmetric spread of all nodes over all locations
559559 - Disadvantage: no diskless SBD feature allowed for all two node clusters
560560* The two node cluster and more than two locations
561- - Advantage: SBD device can be provided from there (must be HA himself )
562- - Advantage: cluster could operate with three SBD devices from different locations
561+ - Advantage: SBD device can be provided from there (must be HA itself )
562+ - Advantage: cluster can operate with three SBD devices from different locations
563563 - Disadvantage: no diskless SBD feature allowed for all two node clusters
564564* The three node cluster and two locations
565565 - Advantage: less complex infrastructure
0 commit comments