iSCSI multipathing gives a Linux host more than one usable route to the same storage device. The host’s device-mapper multipath layer groups the resulting block-device paths and selects among eligible routes; if one fails, it can route I/O over another active path. Understanding the difference between an iSCSI session and a path is key: a session is a protocol association, while a path is a host-visible route to a logical unit.
What is iSCSI multipathing?
iSCSI carries SCSI storage traffic over an IP network. An initiator is the endpoint requesting access to storage; a target provides that storage. Multipathing makes multiple host-visible routes available to the same logical unit, often using more than one host network interface and more than one target portal.
As an Amazon Associate I earn from qualifying purchases.
Linux device-mapper multipath presents the paths as one multipath device to applications and selects a path for I/O according to its configured policy. In simplified form, the stack is: application → multipath virtual device → one of several SCSI block-device paths → iSCSI session and connection → target portal → target storage. The underlying block-device path is the route the multipath layer evaluates; it is not another name for an iSCSI session.
Storage architecture can make the physical route less direct than this diagram suggests. A distributed array may accept a session at one member and forward an I/O internally to another member that serves the requested block. As described in the Linux kernel dm-switch documentation, multiple sessions over multiple host and target interfaces can coexist with this kind of internal forwarding.
#1 Best Overall
- Full-Scale Professional Network-Attached Storage – Business storage solution with hard drives included and optimized to store, share, and back up data for environments of any size.
- Advanced Hardware and Firmware – Product designed for stability and security, capable of handling heavy data loads without dropping performance.
- Purpose-Built for Data Protection – Secure NAS on closed system with 256-bit drive encryption, two-factor authentication, and flexible backup features to keep your data safe.
- Snapshots for Instant Data Backup and Recovery – Snapshots can be created and used to recover data near instantaneously, with little or no system disruptions, and mitigate ransomware.
- Fast Data Transfers – Native 10GbE port for high-speed file transfers with no cable upgrade needed.
How are targets, portals, sessions, connections, and paths different?
The iSCSI protocol distinguishes nodes, sessions, connections, and portals. A target portal is an addressable route to a target; a session is the protocol association between an initiator and a target. A session may include connections, so neither a portal nor a connection should be treated as interchangeable with a session. The protocol terminology is defined in RFC 7143.
A path is the host’s usable route to a logical unit, represented to Linux multipath by an underlying SCSI block device. It can depend on the initiator interface, network, target portal, and storage design. Multiple interfaces and portals often create multiple sessions and block-device paths, but the exact mapping varies by transport and array.
Rank #2
- Full-Scale Professional Network-Attached Storage – Business storage solution with hard drives included and optimized to store, share, and back up data for environments of any size.
- Advanced Hardware and Firmware – Product designed for stability and security, capable of handling heavy data loads without dropping performance.
- Purpose-Built for Data Protection – Secure NAS on closed system with 256-bit drive encryption, two-factor authentication, and flexible backup features to keep your data safe.
- Snapshots for Instant Data Backup and Recovery – Snapshots can be created and used to recover data near instantaneously, with little or no system disruptions, and mitigate ransomware.
- Fast Data Transfers – Native 10GbE port for high-speed file transfers with no cable upgrade needed.
There is also a terminology wrinkle in Linux open-iscsi: iscsiadm uses “node” for a node record associated with a target portal. That usage differs from the RFC’s protocol-level node terminology. See the Debian Bookworm iscsiadm manual when interpreting its records and command modes.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →How does Linux choose a path?
A path selector is a policy for choosing among paths, not a guarantee of a particular performance result. The choice should fit the path groups and priorities exposed by the array, the workload, and the support guidance for the Linux platform and storage system. Two selectors documented by the Linux kernel illustrate the difference:
Rank #3
- Full-Scale Professional Network-Attached Storage – Business storage solution with hard drives included and optimized to store, share, and back up data for environments of any size.
- Advanced Hardware and Firmware – Product designed for stability and security, capable of handling heavy data loads without dropping performance.
- Purpose-Built for Data Protection – Secure NAS on closed system with 256-bit drive encryption, two-factor authentication, and flexible backup features to keep your data safe.
- Snapshots for Instant Data Backup and Recovery – Snapshots can be created and used to recover data near instantaneously, with little or no system disruptions, and mitigate ransomware.
- Fast Data Transfers – Native 10GbE port for high-speed file transfers with no cable upgrade needed.
| Selector | Decision | What to keep in mind |
|---|---|---|
queue-length |
Selects the path with the fewest in-flight I/Os. | It compares current queue depth; it does not estimate service time. See the kernel queue-length documentation. |
service-time |
Estimates service time using the size of I/O already in flight plus incoming I/O, relative to the path’s configured throughput value. | A path with zero relative throughput is not selected while positive-throughput paths are available. See the kernel service-time documentation. |
These mechanisms do not establish that one selector—or round robin—is universally best. Nor do they establish a transferable speedup: actual results depend on the storage and host configuration and workload.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How does multipathing handle a failed route?
When a route becomes unusable, the multipath layer can mark its path failed and direct eligible I/O over other active paths. Device-mapper reports PATH_FAILED and PATH_REINSTATED events; event fields identify the affected path and the valid-path count. The device-mapper uevent documentation describes these notifications.
Rank #4
- Synology DiskStation DS620slim, made for a variety of server roles such as iSCSI targets backup, file storage, email servers, and domain controllers!
- Intel Celeron J3355 Dual-Core 2.0GHz 2MB CPU, Up To 2.5GHz Turbo; 6GB DDR3L Synology SDRAM Memory; 1TB (2 x 500GB) SATA III Solid State Drives for Ultra Fast Storage; 2 x RJ-45 1GbE LAN Port (with Link Aggregation / Failover support)
- 2 x USB 3.0 Port; Btrf File System for Advanced LUN iSCSI Service
- Synology NAS chassis comes in a sealed box.
- Hard drives and memory upgrades included separately NOT installed, installation required.
Failover is not a promise of uninterrupted application behavior. The outcome depends on the host, target, application, and configuration. The documented mechanisms do not define one universal failure-detection or recovery interval, so do not infer a fixed timeout from the presence of multiple paths.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →How can I check iSCSI sessions and paths on Linux?
Check discovery records, logged-in sessions, and block paths as separate layers. A discovered portal or node record does not by itself prove that a session is logged in or that multipath is using the resulting block device.
- Inspect discovery and node records. Use
iscsiadmdiscovery and node modes to review targets and portal-associated records. Its manual documents the available modes and options. - Check logged-in sessions and connections. Use
iscsiadmsession mode to display active sessions and connections; this is distinct from merely having discovery records. - Inspect the multipath device and its component paths. Check the deployed host’s multipath status and confirm which underlying block devices are available and active for the logical unit.
- Use the host’s supported login and rescan procedures if needed.
iscsiadmsupports login and rescanning, but distribution behavior and command details can differ. Follow the deployed distribution and array guidance, then recheck sessions and path status.
The iscsiadm(8) manual notes that discovery can specify multiple interfaces, while each node-mode invocation accepts one interface. Validate the actual sessions and multipath state on the system in question rather than assuming the records map one-to-one.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
