Plugin:PCIe Errors

From collectd Wiki
Jump to: navigation, search
PCIe_Errors plugin
Type: read
Callbacks: init, config, read, shutdown
Status: supported
First version:
Copyright: 2019 Intel Corporation
License: MIT license
Manpage:
List of Plugins

The purpose of this feature is to monitor and report PCI Express errors. There are two mechanisms for error handling. First is base line which is mandatory for every PCIe device, but provides only limited information as there are only four error types. It resides in Device Status register of PCI Express capability. The second is extended capability with Advance Error Reporting. It can provide detailed information about errors set on device. Its occurrence is optional, and not every device provides this extended information.

Synopsis

see PCIe Errors High Level Design

<Plugin pcie_errors>
       Source "sysfs"
       AccessDir "/sys/bus/pci"
       ReportMasked false
       PersistentNotifications false
       FirstFullRead false
       LogFile "/var/log/syslog"
       <MsgPattern "AER">
               <Match>
                       Name "aer error"
                       Regex "AER:.*error received"
                       SubmatchIdx -1
               </Match>
               <Match>
                       Name "incident time"
                       Regex "(... .. ..:..:..) .* pcieport.*AER"
                       IsMandatory false
               </Match>
               <Match>
                       Name "root port"
                       Regex "pcieport (.*): AER:"
               </Match>
               <Match>
                       Name "device"
                       Regex " ([0-9a-fA-F:\\.]*): PCIe Bus Error"
               </Match>
               <Match>
                       Name "severity"
                       Regex "severity=([^,]*)"
               </Match>
               <Match>
                       Name "error type"
                       Regex "type=(.*),"
                       IsMandatory false
               </Match>
               <Match>
                       Name "id"
                       Regex ", id=(.*)"
               </Match>
       </MsgPattern>
</Plugin>


Parameters

needs to be added

Metrics


Metric/Feature/Input

Name

Date Type

Format Example

Description

Dependencies

Limitations

Comments


PCIe AER Plugin

-

-

Plugin to provide PCIe AER metrics, errors, notifications & device information

Depends on sysfs and proc file systems

To be used on little endian systems.


Feature

Device Domain

Hex

10

The PCI address domain consisting of three distinct address spaces: configuration, memory, and I/O space.

None



Feature

Device Bus

Hex

10

PCIe Bus number

None



Feature

Device ID

Hex

3597

PCIe Device ID of the device

None



Feature

Device Function

Hex

10

Bus:Device.Function notation used to succinctly describe PCI and PCIe devices

None



Feature

Instance Type

Text

correctable/uncorrectable

PCIe instance type

None



Feature

Severity

Text

Fatal/Non-fatal

Severity flag indicating nature of severity of uncorrectable errors with fatal or non-fatal error types

None



Feature

Persistent Notification

Text

True/False

If any uncorrectible error is already reported once, persistent flag is set in the plugin and not reported again

None



Metric

Uncorrectable Error

Text

uncorrectable

The errors which don’t have impact on integrity of the PCI Express fabric, but data/information is lost. Non-fatal errors are corrupted transactions that can’t be corrected by PCIe hardware.

However, the PCI Express fabric continues to function correctly and other transactions are unaffected, only particular transaction is affected. Recovery from a non-fatal error may or may not, depends on device-specific software associated with the requester that initiated the transaction

None



Metric

Correctable Error

Text

correctable

the errors which may have an impact on performance (like latency, bandwidth), but no data/information is lost and PCIe fabric remains reliable. Such errors are corrected by hardware and no software intervention is required

None



Metric

Severity Non-Fatal Error

Text

non_fatal

Error severity indicating no reboot necessary

None



Metric

Severity Fatal Error

Text

fatal

Error severity indicating reboot necessary

None



Metric

Unsupported Request

Text

unsupported

This error occurs when an endpoint or a root port recieves any of a set of transactions as defined by PCIe Spec defined in [1]. In all cases the TLP is deleted in the Hard IP block and not presented to the Application Layer. If the TLP is a non-posted request, the Hard IP block generates a completion with Unsupported Request status.

Depends on what's exposed in sysfs and proc file systems



Metric

Data Link Protocol Uncorrected Error

Text

Data Link Protocol

This error occurs when a sequence number specified by the Ack/Nak block in the Data Link Layer (AckNak_Seq_Num) does not correspond to an unacknowledged TLP.

Depends on what's exposed in sysfs and proc file systems



Metric

Surprise Down Uncorrected Error

Text

Surprise Down

When the PCIe device goes down without a notice

Depends on what's exposed in sysfs and proc file systems



Metric

Poisoned TLP Uncorrected Error

Text

Poisoned TLP

anytime a poisoned TLP is destined to PCIe device, IIO module will drop the poisoned data packet, contain the error in the domain that it was detected in, bring down the link, and signal a fatal error to SW/FW

Depends on what's exposed in sysfs and proc file systems



Metric

Flow Control Protocol Uncorrected Error

Text

Flow Control Protocol

An uncorrected error in flow control protocol found in transaction layer that prevents flow control credits transactions being sent. This error occurs when a component does not receive update flow control credits with the 200 µs limit.

Depends on what's exposed in sysfs and proc file systems



Metric

Completion Timeout Uncorrected Error

Text

Completion Timeout

This error occurs when a request originating from the Application Layer does not generate a corresponding completion TLP within the established time. It is the responsibility of the Application Layer logic to provide the completion timeout mechanism. The completion timeout should be reported from the Transaction Layer using the cpl_err[0] signal.

Depends on what's exposed in sysfs and proc file systems



Metric

Completer Abort Uncorrected Error

Text

Completer Abort

The Application Layer reports this error using thecpl_err[2]signal when it aborts receipt of a TLP.

Depends on what's exposed in sysfs and proc file systems



Metric

Unexpected Completion Uncorrected Error

Text

Unexpected Completion

This error is caused by an unexpected completion transaction as listed in [1]. The TLP is not presented to the Application Layer; the Hard IP block deletes it.

Depends on what's exposed in sysfs and proc file systems



Metric

Receiver Overflow Uncorrected Error

Text

Receiver Overflow

This error occurs when a component receives a TLP that violates the FC credits allocated for this type of TLP. In all cases the hard IP block deletes the TLP and it is not presented to the Application Layer.

Depends on what's exposed in sysfs and proc file systems



Metric

Malformed TLP Uncorrected Error

Text

Malformed TLP

This error is caused by an unexpected completion transaction as listed in [1]. The TLP is not presented to the Application Layer; the Hard IP block deletes it.

Depends on what's exposed in sysfs and proc file systems



Metric

ECRC Uncorrected Error Status

Text

ECRC

ECRC ensures end-to-end data integrity for systems that require high reliability. When the ECRC generation option is turned on, errors are detected when receiving TLPs with a bad ECRC. More details in [2]

Depends on what's exposed in sysfs and proc file systems



Metric

Unsupported Uncorrected Error Request

Text

Unsupported

This error is caused by an unexpected completion transaction as listed in [1]. The TLP is not presented to the Application Layer; the Hard IP block deletes it.

Depends on what's exposed in sysfs and proc file systems



Metric

ACS Violation Uncorected Error

Text

ACS Violation

Violation in Access Control Services. More details in [3]

Depends on what's exposed in sysfs and proc file systems



Metric

Internal Uncorrected Error

Text

Internal Uncorrected

An error associated with a PCI Express interface that occurs within a component and which may not be attributable to a packet or event on the PCI Express interface itself or on behalf of transactions initiated on PCI Express. More details in [4]

Depends on what's exposed in sysfs and proc file systems



Metric

MC Blocked TLP Uncorrected Error

Text

MC Blocked TLP

An error with Multicast TLP processing. More details in [5]

Depends on what's exposed in sysfs and proc file systems



Metric

Atomic Egress Blocked Uncorrected Error

Text

Atomic Egress Blocked

Error with setting AtomicOp Egress Blocking bit. More details in [6]

Depends on what's exposed in sysfs and proc file systems



Metric

TLP Prefix Blocked Uncorrected Error

Text

TLP Prefix Blocked

The TLP Prefix mechanism extends the header size by adding DWORDS to the front of headers that carry additional information. The uncorrected error reflects failure in the process. More details in [7]

Depends on what's exposed in sysfs and proc file systems



Metric

Receiver Error Status Corrected Error

Text

Receiver Error Status

Receiver error at PCIe physical layer

Depends on what's exposed in sysfs and proc file systems



Metric

Bad TLP Status Corrected Error

Text

Bad TLP Status

This error occurs when a LCRC verification fails or when a sequence number error occurs.

Depends on what's exposed in sysfs and proc file systems



Metric

Bad DLLP Status Corrected Error

Text

Bad DLLP Status

This error occurs when a CRC verification fails.

Depends on what's exposed in sysfs and proc file systems



Metric

Replay NUM Rollover Corrected Error

Text

Replay NUM Rollover

This error occurs when the replay number rolls over.

Depends on what's exposed in sysfs and proc file systems



Metric

Replay Timer Timeout Corrected Error

Text

Replay Timer Timeout

This error occurs when the replay timer times out

Depends on what's exposed in sysfs and proc file systems



Metric

Advisory Non-Fatal Corrected Error

Text

Advisory Non-Fatal

The error are reported and signaled as ERR_COR, ERR_NONFATAL, ERR_FATAL or not signaled at all, depending upon the role of the agent that detects the error and whether the agent implements AER as an advisory capacity to application. More details in [8]

Depends on what's exposed in sysfs and proc file systems



Metric

Corrected Internal Corrected Error

Text

Corrected Internal

An error associated with a PCI Express interface that occurs within a component and which may not be attributable to a packet or event on the PCI Express interface itself or on behalf of transactions initiated on PCI Express. More details in [4]

Depends on what's exposed in sysfs and proc file systems



Metric

Header Log Overflow Corrected Error

Text

Header Log Overflow

When a header is logged, the header is that of the first TLP that was lost or corrupted by the Uncorrectable Internal Error. More detilas in [9]

Depends on what's exposed in sysfs and proc file systems


Example Graph

None yet. Add one now!

Dependencies

Depends on sysfs and proc file systems

Caveats

History

See also