FAQ

Key terms and definitions are listed in the glossary.

Organizations, vDCs, Pools, Regions

I cannot access vCloud Director though there are funds on my account

Most likely, you are trying to access the Moscow vCloud Director account through the web panel of the Saint Petersburg vCloud Director (or vice versa). If so, please check the web panel address:

  • Moscow: vcd-msk.selectel.ru;
  • Saint Petersburg: vcd.selectel.ru.

vCloud Director account will be blocked for 10 minutes after 5 failed password entry attempts. You can change the password in the Users tab of the Selectel Control panel.

I cannot see my vDCs in the vCloud Director panel. What should I do?

Our infrastructure has 2 vCloud Director panels — for resources in Moscow (vcd-msk.selectel.ru) and Saint Petersburg (vcd.selectel.ru). Each vCloud Director only contains data related to its region’s vDC. Therefore, Moscow resources are managed through the Moscow vCloud Director and Saint Petersburg resources through the Saint Petersburg vCloud Director.

To open the vCloud Director of the corresponding region, click the title to the vDC card in the Control panel.

With Multisite, you can put two regions in a single panel. See more in the documentation.

How do I move the VM from one pool to another?

You can only change the VM pool for the entire vApp:

  1. Move the VM to another vApp.
  2. Disable this vApp in vCloud Director and then select Move to in the context menu.
  3. In Virtual Datacenter section, select the data center with another pool type.
  4. Click OK.

For a Windows-operated VM, submit a request via the ticket system.

For a VM with up to 40 GB of the disk volume, migration will take no longer than 10 minutes.

I need to move the vDC to another account. Is this possible?

Yes, it is possible, but there are some restrictions and limitations:

  • the new account must have no vDCs;
  • you must move all vDCs from the old account to the new one;
  • the migration is Organization-level — i.e. the entire Organization and all vDCs are moved;
  • there is no VM idle time;
  • all the Organization users are saved and also moved.

To order migration, submit a request via the ticket system.

What are the limits on vDC resources and quantity?

Standard vDC resources limits:

Pool vCPU RAM, GB Disk, TB
Gold 100 500 3
Silver 100 500 5
Gold DR 100 500 2

vDCs per account: 5.

At the client’s request, we can lift the limits on vDC resources and quantity.

How do I remove the vDC?

You can only remove empty vDCs. Beforehand, remove all vApps, VMs, vApp templates, and catalogs via vCloud Director.

In the Selectel Control panel, go to the Cloud powered by VMware section. Expand the (⋮) menu in the vDC card and click Delete Datacenter.

Networks

How many networks can I create within one vDC?

One vDC can comprise up to 50 networks. At the client’s request, we can increase that limit.

What about the local network to other services?

vDC networks can be connected into a local network of dedicated servers, colocated servers, stands, and leased network equipment.

You can use any addressing within the local network.

QnQ is not supported. If you use QnQ in the physical local network, you will have to create the new VLAN and connect it to VMware. Submit a request via ticket system. In the request, specify the vDC that will contain the network, addressing in the physical network, and the desired addressing in the VMware network.

What is Edge Gateway (EG)?

EG is a software-based router through which the VM accesses the network.

EG (Edge Gateway) is the technical implementation of the NSX functionality. This drives such mechanisms as DHCP, Load balancing, Firewall, Dynamic Routing and VPN. See the detailed instructions in the official VMware documentation.

When creating the vDC, the Compact-size EG is created automatically (Compact is the mode of the lowest performance and price). You can manage the created EG and adjust its size.

EG prices are provided on the service page.

Is any redundancy provided for Edge Gateway? If so, how does it work?

Yes, Edge Gateway has the option called High Availability (HA). This option is connected upon the request via the ticket system.

The price is calculated as follows: “no-HA EG x 2”.

When the user enables EG HA (not to be confused with vSphere HA — these are different mechanisms), another VM with the same EG is launched. The primary VM operates in primary mode and the secondary will be on standby. Also, the machines will exchange heartbeat messages. If the primary VM does not respond to the heartbeat, the standby VM becomes the primary. Practically, such a configuration prevents EG operation failure in case the virtualization host goes down by ensuring a quick (up to 15 seconds) switch to the standby EG virtualization host.

You can find more additional information and instructions in the VMware documentation.

Is any network equipment redundancy provided?

Yes, it is. All the network equipment — from switches to routers — is fully redundant. The VM will remain available from public and local networks even if any of the network devices fail.

How do I add public IPs?

You can add IPs to existing Edge routers. To do so, submit a request via the ticket system. The price per address is provided on the service page.

Can I assign a public IP directly to the VW, without NAT?

Yes, you can. We will provide the /29 or /28 Direct Connected subset. If needed, you can order the subnet of up to /24 size via the ticket system. You can use any available address within the provided dedicated subnet. The price per address block is provided on the service page.

The address is assigned automatically when the machine is deployed, provided that:

  • the VM has vmware-tools (Windows)/open-vm-tools (*nix) installed;
  • the machine network interface is connected to the relevant network.

VM cannot access its external IP via NAT. What should I do?

Enable NAT reflection (aka NAT hairpinning, NAT loopback). For example:

Act Interface Source Port Destination Port Proto
DNAT WAN 85.119.148.123 80 10.60.0.6 80 TCP
DNAT LAN 85.119.148.123 80 10.60.0.6 80 TCP
SNAT LAN 10.60.0.6 any 85.119.148.123 any any

NAT/Firewall is unavailable. What is wrong?

If the firewall is disabled, NAT rules will not function.

Make sure the firewall is enabled and rules are set to allow the relevant traffic. Otherwise, enable the firewall and add the relevant rules manually.

What firewall is used?

We suggest using a NSX-integrated firewall.

Disks

Virtual disk speed

The table below presents the overall disk performance for all read and write operations by 32-kB block:

Gold pool Silver pool
Fast vSSD — 25000 IOPS Fast vHDD — 2000 IOPS
vSSD — 5000 IOPS vHDD — 150 IOPS

Can I connect virtual disks from different pools to the same VM?

No, you cannot. You can use different types of disks when creating a virtual machine, but only inside the pool where it is created, for example for both vSSD and Fast vSSD or vHDD and Fast vHDD. You can create VMs in different pools, see them in your organization in vCloud Director, and merge them into a single network.

How to use and manage VM Snapshots?

You can create, delete, and restore snapshots in vCloud Director.

One VM is the same as one snapshot. When creating a VM snapshot, they are created for all disks connected to the VM. There are no limitations on disk snapshots.

You cannot create a disk from a snapshot. You can restore the VM version created in a snapshot.

Do not keep the snapshot for more than 72 hours. If you do so, it may lead to deterioration of the VM disk’s performance. For long-term backup, use the backup service.

In which formats can templates be imported?

vCloud Director supports the following template formats: ova and ovf. An ovf template must be imported with the accompanying files, namely vmx, mf, and vmdk.

I cannot select the storage policy when creating a VM from the template in the html5 console

When you create a VM, the last column of the Templates table only shows the template storage.

You can adjust the storage policy after creating the VM:

  1. Click Details in the created VM card.
  2. In the Storage Policy section, select the most preferable one and click Save.

Please note that you can select the storage policy relevant to the vDC pool (vHDD and Fast vHDD for Silver, vSSD and Fast vSSD for Gold).

Backup

How is backup organized?

Backups are organized by the Veeam Enterprise Manager using the Veeam Backup and Replication (VBR) system.

Read more about backup in the documentation.

How do I manage the backup?

Veeam Backup and Replication (VBR) is connected individually for enterprises based in Moscow and Saint Petersburg.

To access Veeam Enterprise Manager, use this link:

  • vbr.selectel.ru/vCloud/<org-name>/ for Saint Petersburg;
  • vbr-msk.selectel.ru/vCloud/<org-name>/ for Moscow,

where <org-name> is the enterprise name.

To log in, enter the credentials of the appropriate-region account.

Miscellaneous

How do I monitor the VM load?

Go to vCloud Director. In the VM card, click Details. Find the last tab, Monitoring Charts. From there, you will see VM load charts for all VM disks.

Metrics are kept for 90 days. The vCloud Director displays the data for the last week, day, hour, and half hour with a step of 1 minute. You can also access older metrics via API: current values, older values.

Available metrics:

  • cpu.usage.average;
  • cpu.usage.maximum;
  • cpu.usagemhz.average;
  • disk.provisioned.latest;
  • disk.used.latest;
  • mem.usage.average;
  • virtualDisk.numberReadAveraged.average;
  • virtualDisk.numberWriteAveraged.average;
  • virtualDisk.read.average;
  • virtualDisk.write.average.

Metrics are captured by means of the hypervisor, so you do not need to install any other software to your VM.

The vCloud Director UI has disk.read.average and disk.write.average e4which are used for classic data storage systems. virtualDisk.read.average and virtualDisk.write.average metrics are used for disks of VMs located in vSAN.

Does Cloud powered by VMware support nested virtualization?

Yes, but we cannot guarantee adequate operating performance. Read more in the VMware documentation.

Max CPU, RAM, Virtual Machine disk volumes

For the performance-guaranteed VM, max properties are as follows:

  • vCPU — 36;
  • vRAM — 256 GB;
  • virtual disk — 10 TB.

Is Microsoft software supported by the Cloud powered by VMware?

Yes, it is. You can create a VM from the standard templates of Windows Server editions:

  • Windows Server 2019 Standard;
  • Windows Server 2016 Standard;
  • Windows Server 2012R2 Standard.

You can use the server version with the pre-installed Desktop Experience package. License billing:

  • license price per core: ~RUB 277/month or 38 kopecks per hour;
  • hourly rate only applies to active Windows-based VMs;
  • if the VM was activated for 10 minutes within an hour, the license will be billed for the full hour;
  • the VM license cannot be revoked — you can only delete the VM itself.

If you moved a VM or created one with the Windows Server from your image, please let us know. We will enable the license for such a VM manually.

Can Selectel engineers access the customer vDC or data?

Only a few Selectel specialists can access the client’s infrastructure and data. This access must fulfil our obligations regarding the infrastructure performance support we provide to the customer.

As per regulations, customer data cannot be copied, modified, nor removed without the customer’s explicit permission.

How are customer clouds isolated from each other? If a customer loads the host, will it affect their digital neighbors?

In vSphere, there is a tool called DRS (Distributed Resource Scheduler). This tool automatically and seamlessly moves neighbors to less loaded hosts.

How could I move my VMs to the Selectel infrastructure?

You can use vCloud Availability Appliance to move your virtual machines. Read more about configuring the client in the “Setting up the on-prem vCloud Availability” article.

Alternatively, you can export/import your VMs through the OVA template:

  1. To export the machines, move them to the individual vApp. Expand the menu in the vApp card and click Download. You will download the OVA template.
  2. Then import this template to vCloud Director (Selectel infrastructure-based instrument). Use the guide How to Import a Virtual Machine.

Is Terraform support available?

Yes, you can find more about it here: https://www.terraform.io/docs/providers/vcd/index.html

What APIs are available?

We recommend our customers use vCloud Director API to control VMs, networks, disks, and other resources. Visit the VMware website to see relevant documentation:

We use vCloud Director 9.7. The service supports API versions from 20.0 to 32.0.

We will provide the individual Selectel API for organizing the virtual data centers. Currently, this API and related documentation is under development.

How can I make sure that VMs are always on different hosts in the cluster?

You can create Affinity rules to set VM host rules for VMware DRS in vCloud Director.

There are two types of rules:

  • Affinity rules instruct the DRS to host VMs on a single virtualization host;
  • ⦁ Anti-Affinity rules allow you to host VMs on different cluster hosts, which increases the availability of the service when one or more hosts fail.

These rules vary according to the type:

  • hard (required) rules do not allow the rule to be ignored in the absence of the required number of available hosts. For example, the rule specifies 10 VMs, but only 8 hosts are available in the cluster (in this case, the machines will not start);
  • soft (preferred) rules allow you to ignore the rule’s requirements if there are no resources that meet the rule.

To configure Anti Affinity Rules:

  1. Go to the vDC management section.
  2. In the Compute section, select Affinity Rules.
  3. Select New in the required section (Affinity Rules or Anti-Affinity Rules).
  4. Specify a rule name.
  5. Specify the rule type using the Required checkbox — the installed one will create a hard rule, the removed one will create a soft rule.
  6. Click Save.

We do not recommend creating rules that include many VMs. It would be better to divide VMs into service groups for which they are responsible and then create separate rules for each VM group.