FC Pools on a Managed Array (COMSTAR)

Use this page when Karios should create and manage the disks on your Fibre Channel SAN array. Karios logs in to the array, carves one volume for every VM disk, and shows it to the host that runs the VM.

If you want to attach LUNs that already exist on your own FC array, use Generic FC Pools (Existing LUNs) instead. For shared concepts (ports, paths, readiness, health), see Fibre Channel Storage.

UI path: Control Center -> Storage -> Instance Storage

When to Use

  • You have a COMSTAR array (illumos or a compatible ZFS appliance) with FC target ports on the fabric.

  • You want VM root disks, data disks, and templates to live on that array.

  • You want Karios to create, grow, and remove the array volumes for you.

What Karios Does on the Array

Once a pool is in use, Karios manages these objects on the array:

  • One FC target group named karios-fc. The array ports you set up join it.

  • One host group per Karios host, named kh-<host name>. Each host is admitted by its own initiator WWPN.

  • For each VM disk: one ZFS volume under the dataset you name, one logical unit, and one view that shows it to the host group of the host running the VM.

Karios does not change the switch. Zoning stays yours.

Before You Start

  • [ ] You are a platform administrator. Only a platform administrator can register an array or set up one of its ports. Creating a pool on a registered array needs the storage manage permission.

  • [ ] The array’s STMF service is online and its FC ports run in target mode.

  • [ ] The ZFS pool that will hold the volumes exists, for example sanpool. Pick a dataset path under it, for example sanpool/karios. The dataset itself can be new.

  • [ ] You have a management address, a username, and a password for an SSH account that can run the array’s COMSTAR and ZFS commands, and the array’s SSH host key from ssh-keyscan -t ecdsa <management address>.

  • [ ] Use array ports that are dedicated to Karios. A port must join karios-fc, and joining takes the port offline for a moment. Karios refuses to join a port that something outside Karios is logged in to.

  • [ ] The fabric meets the checklist in Fibre Channel Storage: every host in scope has an HBA and is zoned to the array’s ports.

Important

FC has no CHAP. Hosts are admitted by per-host WWPN masking on the array and by the switch’s zoning. Keep the fabric dedicated to storage.

Task 1: Register the Array

When to Use: Do this once per array, before you create the first pool on it.

Purpose: Store the array’s address, credentials, and host key so Karios can manage it.

Steps:

  1. Open Control Center -> Storage -> Instance Storage.

  2. Click Add Instance Storage. The Add New Instance Storage panel opens.

  3. Set Protocol to fc. In Driver, pick comstar (carves volumes on an array). Only drivers that carve volumes and support FC are offered here.

  4. If no array is registered yet, the Register array form is already open. Otherwise, click Register array next to Array.

  5. Fill in Array name, Driver (comstar), Dataset, Management address, Username, Password, and the Array SSH host key. The fields are the same as for an iSCSI array; see iSCSI Pools on a Managed Array.

  6. Click Register. Karios probes the array before it stores it.

Expected Outcome:

  • The array appears in the Array list and is selected.

  • The panel shows SSH host key verified: and the key.

  • The Array FC ports card fills with the array’s target ports.

If this fails:

  1. KF-5010: the array did not answer. Check the management address, the network path from the management node, and the credentials.

  2. KF-5011: the array answered but the probe failed. Check the credentials, that the array runs COMSTAR with STMF online, and that the ZFS pool named in Dataset exists.

  3. The array lists no FC target port: its HBA ports are not in target mode. Fix that on the array, then click Re-check.

Task 2: Set Up the Array Ports

When to Use: Do this once per array port, before the first pool on the array.

Purpose: Add each port to the karios-fc target group so it serves this array’s LUNs.

Steps:

  1. In Add New Instance Storage, with the array selected, read the Array FC ports card. Each row shows the port’s WWPN and a Status:

    • Serving: online and already in the target group.

    • Needs setup: online but not in the target group yet.

    • In use elsewhere: online but something outside Karios is logged in to it.

    • Offline: link down. Check the cable, SFP, and switch port.

  2. Click Set up on each Needs setup port you want Karios to use. Read the confirmation: the port goes offline for a moment while it is added. If the array could not count the sessions on the port, the dialog says so; check on the array first.

  3. Click Set up in the dialog.

Expected Outcome:

  • The port reads Serving.

  • On the Hosts card, hosts zoned to that port now read Ready and count it.

If this fails:

  1. The port was not set up: the array refused the join, usually because the port carries sessions from outside Karios. Move them off the port or pick another port.

  2. Admin only: you are not a platform administrator. Ask one to set up the port.

Task 3: Create the Pool

When to Use: Do this after the array is registered and at least one port is serving.

Purpose: Create an FC pool that places VM disks on the array.

Steps:

  1. In Add New Instance Storage, set Scope:

    • Zone: every cluster in the zone can use the pool.

    • Cluster: one cluster only. Pick the Pod and Cluster.

  2. Pick the Zone and enter a Name for the pool.

  3. Set Protocol to fc, Driver to comstar, and pick the registered array.

  4. Under Ports to watch for health, leave every serving port ticked, or untick the ports this pool’s health should not count. At least one stays ticked. The pool is served through every port the array has set up either way.

  5. Read the Hosts card. Every host in scope should read Ready and name the ports it sees. For a host that reads Not zoned, click Show zoning commands, enter the active zoneset and VSAN, and paste the commands on the switch. Then click Re-check. For a host that reads Unknown, wait for its report, or restart the Karios host service on it. Create stays off while no host in scope reaches the array.

  6. Check Minimum paths. Karios proposes the value from what the hosts see, capped at the ticked ports. If it reads 1 and a host has measured one path, tick the single path acknowledgement.

  7. Optional: add Storage tags to steer disks of certain volume profiles to this pool.

  8. Click Add Instance Storage.

Expected Outcome:

  • The pool appears in the list with protocol FC COMSTAR.

  • On the pool’s detail page, the Fabric card shows Health as Up once every host in scope has reported. With some hosts not zoned yet it reads Degraded and names them.

If this fails:

  1. KF-4013: a host sees fewer paths than Minimum paths. The message names the host, its WWPN, and how many of the pool’s ports it is logged in on. Zone the host to the missing port, bring the port back, or lower Minimum paths.

  2. The create is refused because no host in scope reaches the array: zone at least one host, or pick a scope that holds a zoned host.

  3. Set up at least one port before creating the pool.: go back to Task 2.

  4. The button goes off after you fill the form: a host in scope has not reported yet, so the form has nothing to count. Restart the Karios host service on that host, click Re-check, and try again.

Fabric Card Reference

Open the pool from the list to see its Fabric card.

Row

What it shows

Health

Pool health from the hosts. See Fibre Channel Storage.

Registration

Whether the platform has accepted the pool.

Array and Dataset

The registered array and the dataset volumes are carved under.

Protocol

FC COMSTAR.

Minimum paths

The pool’s path floor. A Single path chip appears when the pool or a host has only one path.

Ports to watch

The array ports health is measured against, with Edit. Saving re-checks the ports against the array; a port not yet in the target group is joined, which takes it offline for a moment. Only an administrator can save.

Admin switch

The pool’s admin state, for example Up, Disabled, or Maintenance.

Hosts

Each host’s reachability: how many of the pool’s ports it sees, when it last reported, and any disk on it that is not healthy.

Last error

The most recent error, if any.

Task 4: Work With Disks on the Pool

When to Use: Use this when you create, grow, or remove VM disks that land on the pool.

Purpose: Know what Karios does on the array for each disk action.

Steps:

  1. Create VMs and disks as usual. See Block Storage. Disks land on the pool when its scope and storage tags match the volume profile.

  2. When a disk is first attached to a VM, Karios creates its volume on the array and shows it to that VM’s host. The host finds the new LUN on its HBAs and builds one multipath device for it. Each detach removes the view again.

  3. To grow a disk, use Update Disk Size in Block Storage. Karios grows the volume on the array, the host re-reads the paths, and a running VM sees the new size. Disks can only grow.

  4. When you deploy VMs from a template, Karios keeps one copy of the template on the pool and clones each root disk from it on the array.

Expected Outcome:

  • On the pool’s Fabric card, the Hosts rows name no disk as unhealthy.

  • In Block Storage, the Fabric paths panel reads Healthy for a disk held by the node that answers the request, with one path per array port.

If this fails:

  1. KF-4013 on attach: the host sees fewer of the pool’s ports than the minimum. Check the Hosts row for that host and the port state on the array and switch.

  2. KF-4092: the template copy on the pool still has disks cloned from it. Delete those disks first.

  3. KF-5012: the array refused the operation. Check space and state on the array.

  4. The VM fails to start with Unable to create a deployment: the platform could not place it. Read the Hosts rows; the usual cause is a host short of paths.

Task 5: Disable, Maintain, Change Scope, or Delete the Pool

The lifecycle actions are the same as for an iSCSI pool on a managed array. Follow Task 4 in iSCSI Pools on a Managed Array (COMSTAR): disable to stop new disks, enable maintenance for a window, change scope from Disabled, and delete only from Maintenance once every disk on the pool is gone.

Deleting the pool removes its views and cached template copies from the array. The karios-fc target group, its ports, and the host groups stay on the array; delete them on the array when the last Karios pool is gone.

Limits

  • No CHAP. Zoning and masking decide access.

  • No volume snapshots (KF-4016).

  • No Storage QoS (KF-4016).

  • Disks can grow but not shrink.

  • A host without an FC HBA never holds the pool.

Troubleshooting

Code or message

What it means

What to do

KF-5010 / KF-5011

The array did not answer, or answered but the probe failed.

Check the address, credentials, STMF, and the dataset’s ZFS pool.

KF-5012

The array refused the operation.

Check space and state on the array.

KF-4013

A host sees fewer of the pool’s ports than Minimum paths.

Zone the host to the missing port, bring the port back, or lower the floor.

Not zoned

The host’s HBA is up but sees none of the array’s ports.

Use Show zoning commands and apply them on the switch.

Unknown on a host

The host has not reported yet, or not for three minutes.

Wait a minute, or restart the Karios host service on it. Check that the host service is running.

No FC HBA

The host has no FC adapter.

Scope the pool to a cluster whose hosts have HBAs, or accept Degraded.

In use elsewhere on a port

Something outside Karios is logged in to the port.

Move it off the port, or use another port.

Placement: Disabled by the fabric

No host could reach the pool for about five minutes.

Fix the fabric or the array. Placement comes back by itself.

Unable to create a deployment on VM start

The platform could not place the VM.

Read the pool’s Hosts rows; the management log carries the KF- cause.

Done Criteria

  • [ ] The array is registered with a verified SSH host key and at least two ports read Serving.

  • [ ] Every host in scope reads Ready and the pool reads Up.

  • [ ] A test disk on a running VM shows Healthy with one path per array port.

  • [ ] Your team knows that snapshots and CHAP are not available on this pool.