How Node Eviction Happens in RAC? What is Fencing in RAC? Network and Disk Heartbeat failure

Hello Friends,
Here with one more topic on.
Fencing Basics - Node Eviction in Oracle RAC || Real Application Cluster:
=============================================================
Why are nodes evicted?
• Evicting (fencing) nodes is a preventive measure (it’s a good thing)!
• Nodes are evicted to prevent consequences of a split brain:
- Shared data must not be written by independently operating nodes
- The easiest way to prevent this is to forcibly remove a node from the cluster
How are nodes evicted? - STONITH
EXAMPLE: Network heartbeat failure
What happens, if CSSD is stuck?
How can nodes be evicted?
EXAMPLE: IPMI based eviction on heartbeat failure
Which node gets evicted?
Cluster members can escalate a kill request
Re-Bootless Node Fencing
EXCEPTIONS
• With Oracle Clusterware 11.2.0.2, re-boots will be seen less, unless…:
- IF the check for a successful kill of the IO processes fails → reboot
- IF CSSD gets killed during the operation → reboot
- IF cssdmonitor (oprocd replacement) is not scheduled → reboot
- IF the stack cannot be shutdown in “short_disk_timeout”-seconds → reboot
#Fencing #NodeEviction #exadata #oracle #oracledatabase #oracledba #rac #database #databaseadministration #databases
Please do follow me and support me on,
LinkedIn: / mallik034
TH-cam: / mallik034
Fakebook: / mallik034
Blog: mallik034.blog...
twitter: / mallik034
Instagram: / mallik034
FBPage: / mallik034oracledba
Skype: malluramadurg
Regards,
Mallikarjun Ramadurg
Mobile: [phone]
WhatsApp: [phone]
Email: [e-mail]
LinkedIn: / mallik034
TH-cam: / mallik034
Fakebook: / mallik034
Blog: mallik034.blog...

[@pratik2720]

Great class

[@syedsami6416]

Awesome

[@worldvswords7653]

how the voting disk will determine to evict which node when both the nodes will not able to ping each other through the interconnect but can accees the voting disks.

[@madhu4641]

Hi Mallik, Need clarification, if one node is unable to communicate with Private n/w and as well as Vote disk and how the kill request will send to other node. May be am wrong. which process actually does all this activity

[@madhu4641]

I understood.

[@Mallik034]

Not an issues. In case any concerns or doubt, Please ping me.

[@mughees52]

In production I saw every time when there is a heartbeat failure only node2 gets killed, why is this so, why not node1 gets killed? how is this decided to kill node2??

[@Mallik034]

Check nodes logs and also check the network configuration on node2.
If possible lets start one webex sessions we will check node configuration.
Please ping me on my whatsapp. We will troubleshoot this.

[@arjunm5169]

May be node1 acts an master node because of that node 2 is getting killed.

[@pravatkumarnayak6127]

if node 1 is not able to communicate to node 2., then how come it is possible for node 1 to send kill request to node 2?

[@Mallik034]

It will happen with the help of GRD.

[@SunilKumar-si1qc]

@@Mallik034 if both the nodes updating voting disk highest number , but interconnect is not working, then which node will be evicted

[@zaheerkothwal1491]

Your voice is very low ...