[DOCS-7428] vmmemctl balloon driver update Production Notes Created: 15/Mar/16  Updated: 11/Jan/17  Resolved: 15/Mar/16

Status: Closed
Project: Documentation
Component/s: manual
Affects Version/s: None
Fix Version/s: 01112017-cleanup

Type: Task Priority: Major - P3
Reporter: Byron Grogan Assignee: Allison Reinheimer Moore
Resolution: Duplicate Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Participants:
Days since reply: 7 years, 48 weeks, 1 day ago

 Description   

Hello.

I am creating this ticket on behalf of ricardo.lorenzo, where an issue had been created using Rietveld to the production-notes.txt file.

ricardo.lorenzo 2016/01/26 15:53:42
The memory reclamation when the hypervisor is overcommitting caused the memory
to be ballooned from a VM using a balloon driver residing inside the guest OS.
When the balloon driver expands, it may induce the guest OS to reclaim memory
from guest applications.

Disabling the balloon driver can cause the hypervisor to use its swap as there
is no other mechanism to perform the memory reclamation. Instead of doing that,
we suggest to map and reserve the full amount of memory for the VM running
MongoDB in order to avoid the balloon to be inflated in the local OS if there is
a memory pressure in the hypervisor due to an overcommitted configuration.

The CPU reservation also shouldn't exceed more than 2 virtual CPU per physical
core.

We also recommend disabling the VM live relocation configuring proper exceptions
in the Distributed Resource Scheduler. The live motion of the VM can cause
different performance problems and it will be difficult for the HA mechanisms to
be triggered as the VM is not in a faulty stage.

It is possible to clone a virtual machine running MongoDB.
You might use this function to spin up a new virtual host to add as a member of
a replica set. If you clone a VM with journaling enabled, the clone snapshot
will be valid. If not using journaling, first stop :program:`mongod`, then clone
the VM, and finally, restart :program:`mongod`.

KVM
```

The same recommendation from VMware applies for Memory and CPU overcommitment.


Generated at Thu Feb 08 07:54:15 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.