RHEV 4.2: How to run commands via vdsm client


On RHV 4.2 can be useful to run commands via vdsm, especially during the troubleshooting. In my case I have a cluster where one hypervisor is in “Non Operational Status” due to a CPU missing flag inside the cluster.
On old releases I could use vdsClient command but now it has been replaced. It changed name and syntax:ย  now I can use vdsm-client command.

For example, How can I check CPU flags among the hypervisors? Easy to do!

# vdsm-client Host getCapabilities | grep -i flags
"cpuFlags": "fpu,vme,de,pse,tsc,msr,pae,mce,cx8,apic,sep,mtrr,pge,mca,cmov,pat,pse36,clflush,dts,acpi,mmx,fxsr,sse,sse2,ss,ht,tm,pbe,syscall,nx,pdpe1gb,rdtscp,lm,constant_tsc,arch_perfmon,pebs,bts,rep_good,nopl,xtopology,nonstop_tsc,aperfmperf,eagerfpu,pni,pclmulqdq,dtes64,monitor,ds_cpl,vmx,smx,est,tm2,ssse3,sdbg,fma,cx16,xtpr,pdcm,pcid,dca,sse4_1,sse4_2,x2apic,movbe,popcnt,tsc_deadline_timer,aes,xsave,avx,f16c,rdrand,lahf_lm,abm,3dnowprefetch,epb,cat_l3,cdp_l3,intel_pt,ibrs,ibpb,stibp,tpr_shadow,vnmi,flexpriority,ept,vpid,fsgsbase,tsc_adjust,bmi1,hle,avx2,smep,bmi2,erms,invpcid,rtm,cqm,rdt_a,rdseed,adx,smap,xsaveopt,cqm_llc,cqm_occup_llc,cqm_mbm_total,cqm_mbm_local,dtherm,ida,arat,pln,pts,spec_ctrl,intel_stibp,model_n270,model_Broadwell-IBRS,model_coreduo,model_SandyBridge-IBRS,model_Nehalem,model_Haswell-noTSX,model_Westmere-IBRS,model_Broadwell-noTSX,model_Haswell-noTSX-IBRS,model_Nehalem-IBRS,model_SandyBridge,model_core2duo,model_IvyBridge,model_Penryn,model_IvyBridge-IBRS,model_Westmere,model_Haswell-IBRS,model_Broadwell-noTSX-IBRS,model_Haswell,model_Conroe,model_Broadwell",

Let’s compare this ouput with the one from other hypervisors ๐Ÿ˜€

Advertisements

RHEV 4.2 metrics !!


From 4.2 release is possible to collect RHEV events and metrics from kibana.
The new metric store is not native, it must be manually installed. In my case I created a new guest inside RHEV and on top of it I installed the metric store. Basically it contains Elastic Search/Kibana running as container inside Openshift. Yes, you read well, Openshift! So, you will have a small Openshift running containing multiple containers. This is the installation guide

Following one screenshot of kibana examples dashboard:

RHEV 4.2 – Rhevm shell


From 4.2 release of Redhat Virtualization, rhevm shell is deprecated. You can still have the command but don’t use it because it “does not know” the new objects with new API version.
So, how can u deal and work about automation? The answer is: ansible!
This is a big opportunity for have fun using ansible. Where can you begin?
From the first, simple playbook:

- hosts: localhost
connection: local

tasks:
- name: Obtain SSO token
ovirt_auth:
url: "https://rhevm_url/ovirt-engine/api"
username: "admin@internal"
password: "insert_the_password"
insecure: "true"

- name: List vms
ovirt_vms_facts:
pattern: cluster=Default
auth: "{{ ovirt_auth }}"

- debug:
var: ovirt_vms

You have first to install ovirt-ansible-roles package, then you can run this playbook
It does not use external variables, it connects to rhevm via insecure mode, so this is the simpliest playbook you can use from rhevm to understand how it works.
This playbook will return each detail about vms running on rhevm. It’s extremely verbose but as I said this is just a start point ๐Ÿ™‚

Have fun!

RHEV 4.2: How to check hosted-engine status via cli


Hosted engine is a special virtual machine inside RHEV architecture.
It has a dedicated command, hosted-engine.

How can I check via cli on which host does have the engine is running?
How can I check which hosts are valid candidates for hosting the engine?

The answer is:

[root@dcs-kvm01 ~]# hosted-engine --vm-status

--== Host 1 status ==--

conf_on_shared_storage : True
Status up-to-date : True
Hostname : dcs-kvm01.net
Host ID : 1
Engine status : {"reason": "vm not running on this host", "health": "bad", "vm": "down", "detail": "unknown"}
Score : 3400
stopped : False
Local maintenance : False
crc32 : caa0ce0d
local_conf_timestamp : 353495
Host timestamp : 353495
Extra metadata (valid at timestamp):
metadata_parse_version=1
metadata_feature_version=1
timestamp=353495 (Fri Aug 10 16:34:24 2018)
host-id=1
score=3400
vm_conf_refresh_time=353495 (Fri Aug 10 16:34:25 2018)
conf_on_shared_storage=True
maintenance=False
state=EngineDown
stopped=False

--== Host 2 status ==--

conf_on_shared_storage : True
Status up-to-date : True
Hostname : dcs-kvm02.net
Host ID : 2
Engine status : {"health": "good", "vm": "up", "detail": "Up"}
Score : 3400
stopped : False
Local maintenance : False
crc32 : b1f91d1e
local_conf_timestamp : 351112
Host timestamp : 351112
Extra metadata (valid at timestamp):
metadata_parse_version=1
metadata_feature_version=1
timestamp=351112 (Fri Aug 10 16:34:20 2018)
host-id=2
score=3400
vm_conf_refresh_time=351112 (Fri Aug 10 16:34:20 2018)
conf_on_shared_storage=True
maintenance=False
state=EngineUp
stopped=False

In my case the hosted engine is running on top of dcs-kvm02 host. Another useful information is the Score: each host has different metrics ( ping to default gw, filesystems, cpu load, etc.etc. ) and each metric has a score. Summing each score u can reach the best result, 3400. This score means the host is perfect for host the engine.
In this case you can also understand that this cluster is composited by 2 hypervisors.

First RHEV 4.2 up & running!


Yesterday I installed and configured RHEV 4.2 new infrastructure, I was so funny!

I installed a couple of hypervisors on DELL M630 blades and rhevm via hosted engine. I saw the new deploy via ansible, it’s more simple and fast than the last one. I also appreciated the cleanup utility after a hosted engine deploy failed. I remember in 4.0, in the same condition I had to manually look for written files and to clean them. It was painful to deal with vdsm configuration files.

I also saw ansible adding the second hypervisor, wow!!
It seems more stable and robust than the previous minor releases. Next step in the project will be to move virtual machines from Vmware and make multiple stress tests.

I was surprised connecting for the first time to the new dashboard, it remembers me Cloudforms ๐Ÿ™‚

This is the new dashboard

RHEV vProtect backup


Today I read on linkedin about this project letting customers to have a software backup planned and designed for RHV/Ovirt platfoms.

From the description I read:

#vProtect 3.4+ (code named MARS) delivers even more for the efficient and scalable RHV backup and recovery. Now vProtect also supports RHV/oVirt API v4 and introduces a new backup mode, eliminating the need to use export storage domain.

Marcin Kubacki, Storware and @Jacek Skรณrzyล„ski, Red Hat will discussed in details:

โ€“ RHV backup/restore via export storage domain
โ€“ RHV backup/restore with VM proxy/disk attachment
โ€“ Mount backups and file-level restore
โ€“ Integration with 3rd party backup providers, such as IBM #SpectrumProtect, Dell EMC #NetWorker, #DataDomain, Veritas Technologies LLC #NetBackup and Amazon Web Services S3
โ€“ Orchestration with Open #API

Can’t wait to play with vProtect!

RHEV 4.2 BETA


Can’t wait for play with RHEV 4.2 !
With 4x I began to work from 4.0 release and it was only pain, bugs, bugs, and again bugs. No performance graphics. After running yum update everything went bad.

I’m reading about 4.2 features and I hope now RHEV is more stable and it can gives at least basic graphic informations about the guests. I wait the official “stable” status and I will begin to play with.

Here you can find a description about the main new features.
Let’s have fun ๐Ÿ™‚