- play_arrow Overview
- play_arrow MPC-specific
- play_arrow Power Management
- play_arrow Environment Monitoring
- play_arrow Managing Errors and Alarms
- play_arrow Network Services Mode
- play_arrow Packet Scheduling Mode
- play_arrow Configuration Statements and Operational Commands
Fabric Management on PTX10K devices
This topic covers fabric management features of PTX10K series devices.
Switch Interface Boards (SIBs) create the switch fabric for the PTX10K series devices. There are two models of supported SIBs: the JNP10008 and the JNP10008-SF5. The following table provides information about SIB and line card support on PTX10K devices:
Line Cards Supported | Switch Interface Boards Supported | Packet Forwarding Engines (PFEs) | Fabric Planes | Fabric Redundancy |
---|---|---|---|---|
PTX10K-LC1201-36CD | JNP10008 | 9 PFE | 36 | Not supported |
PTX10K-LC1202-36MR | JNP10008 | 4 PFE | 36 | Yes (5+1 or 4+2) |
PTX10K-LC1301-36DD | JNP10008-SF5 | 4 PFE | 18 | Not supported |
The major fabric management features include, but are not limited to:
Fabric hardening: Fabric hardening is a resiliency feature to detect fabric blackholing and attempt automatic recovery process to restore the Packet Forwarding Engines from blackhole condition.
Fabric fault detection and management: When a fabric component fails, the system reports an error with its severity. You can use the
show system errors active detail
command to view the logged errors. For major fabric errors, you can initiate manual recovery from CLI, or the system removes fabric traffic from the board. For fatal errors, the system attempts recovery by restarting the SIB, with a limit of 3 restart attempts if errors recur.Fabric autoheal: Fabric autoheal is a mechanism that attempts to recover faulty fabric links from a link error condition. Autoheal, attempted for both runtime and initialization time failures, involves bringing down the faulty fabric link and then training it. Junos OS Evolved attempts to recover a faulty link from a maximum of three link error instances (per link) within a span of 24 hours. The autoheal feature is enabled by default. You can use the existing
show chassis fabric errors autoheal
command to view the details of the autoheal actions performed by the software.Note: If you remove a line card or switch fabric card ungracefully and the FRU on the other side of the link reports a link fault, the software attempts to automatically heal the faulty link. Subsequently, it marks the autoheal status as unsuccessful.