Hi,

I love mangassi interface on web desktop amd mobile.
Could anyone share what framework / js is being use here?

ie. Angular / react / any js? Especialy the new post and side bar for mobile

How about the backend? Express ? Monggo db?

Thanks

@stacksofplates thanks a bunch! it shows now!

yum install cockpit-networkmanager
yum install cockpit-storaged

@scottalanmiller about this, so is it correct to say i should use lvm for all array and partition. And use xfs as file system? Avoid ext4 at all cost?

@francesco-provino

Am i right to say better to just use the whole array on the host and locate it to root partition? and use file image instead for flexibility?

Hi All,

Please share you most preferred KVM Management tools here.
Im thinking to use Proxmox. Any thought?

Thanks.

@fuznutz04 said in KVM on Fedora 26 Server edition:

openssh-askpass

try : sudo yum install openssh-askpass

Hi,

Anyone here know what kind of database engine & schema that salesforce use to facilitate multi-tenant SAAS and each tenant has its own custom field that they can define?

I read online, they use SQL (oracle) with EAV (Entity Attribute Value) for custom field schema.
Can anyone confirm this?

is EAV preferred compared to json native column that recently supported by mysql / maria in term of performance?
What is the down side for json column? Lack of join syntax?

Thanks

@jaredbusch
I got project in CRM SAAS development.
So they want to add custom field feature like saleforce.

I am not trying to match their scale, just want to get some information about the database engine and structure, so we can at least stay under best practice guideline for such case.

The options I have currently :

1. EAV
2. JSON Column (Mysql / Maria recently supported this). / POSTGRESQL
3. NOSQL

Just not sure which is the best for my case above.

I got project related to above questions and want to decide the best practice for above case.

Hi,

Anyone here know what kind of database engine & schema that salesforce use to facilitate multi-tenant SAAS and each tenant has its own custom field that they can define?

I read online, they use SQL (oracle) with EAV (Entity Attribute Value) for custom field schema.
Can anyone confirm this?

is EAV preferred compared to json native column that recently supported by mysql / maria in term of performance?
What is the down side for json column? Lack of join syntax?

Thanks

@dafyre may i know why u change from ur owned hosting to theirs?

@fateknollogee yup CentOs normally has better wide range of support. Thats the reason i stick to it even it is less modern than fedora.

@fuznutz04 said in KVM on Fedora 26 Server edition:

openssh-askpass

try : sudo yum install openssh-askpass

@tim_g another one is because Dell always issue RHEL package which normally is compatible with centos. Never try it with fedora, so not sure it will run well. ie. Omsa dell.

Related to KVM, what fedora can offer that CentOs cant?

@tim_g said in Best practice partition & LVM for KVM:

@kuyaz said in Best practice partition & LVM for KVM:

LVG : vg_ssd_critical_vm
/vm (ssd)
LVG : vg_sata_non_critical_vm
/boot 2GB
/bootBIOS 1GB
/root (ALL remaining space)
/swap (32GB)

I'd use XFS. Don't use EXT4. You're correct with using LVM.

I'd go like this:

LVG : vg_ssd_critical_vm
/DATAssd

LVG : vg_sata_non_critical_vm
/DATAhdd (xfs)
/boot 2GB (xfs)
/bootBIOS 1GB
/root 50GB (xfs)
/swap 32GB (swap)
/home whatever (xfs)

As @travisdh1 said, you'll need space for snapshots of the LVMs, but don't do it for the VMs. Use VM snapshots for those instead.

You're VMs should be RAW (.img) for better performance.

So i put the vmware img in the dataHDD?
Please note this is for production server running sql and web apps (it it matters).

What is the disadvantage if i dont separate the datahdd and home from root? Mine is all lumped in root.

Will redo since im not too far off.

Hi all,

If I have :

• raid 1 4TB Sata 7200 rpm
• raid 1 1TB SSD

What is the best practice partition and mount point for above configuration?
I want to utilize ssd for VM that required high IOPs and the rest will be on SATA.

Is it ok if i do like this :

LVG : vg_ssd_critical_vm
/vm (ssd)

LVG : vg_sata_non_critical_vm
/boot 2GB
/bootBIOS 1GB
/root (ALL remaining space)
/swap (32GB)

I plan to install centos with KVM and ovirt on Sata Raid. Should i also separate the host on different partition / LV?

Do I need to spare some free space for LVM snapshot?

All guru please advise... thanks

Hi All,

Please share you most preferred KVM Management tools here.
Im thinking to use Proxmox. Any thought?

Thanks.