| Message ID | 1609969748-17684-1-git-send-email-muneendra.kumar@broadcom.com |
|---|---|
| Headers | show |
| Series | scsi: Support to handle Intermittent errors | expand |
Muneendra, > This patch adds a support to prevent retries of all the io's after an > abort succeeds on a particular device when transport connectivity to > the device is encountering intermittent errors. Applied to 5.12/scsi-staging, thanks! -- Martin K. Petersen Oracle Linux Engineering
On Thu, 7 Jan 2021 03:19:03 +0530, Muneendra wrote: > This patch adds a support to prevent retries of all the > io's after an abort succeeds on a particular device when transport > connectivity to the device is encountering intermittent errors. > > Intermittent connectivity is a condition that can be detected by transport > fabric notifications. A service can monitor the ELS notifications and > take action on all the outstanding io's of a scsi device at that instant. > > [...] Applied to 5.12/scsi-queue, thanks! [1/5] scsi: Added a new error code DID_TRANSPORT_MARGINAL in scsi.h https://git.kernel.org/mkp/scsi/c/962c8dcdd5fa [2/5] scsi: No retries on abort success https://git.kernel.org/mkp/scsi/c/60bee27ba2df [3/5] scsi_transport_fc: Added a new rport state FC_PORTSTATE_MARGINAL https://git.kernel.org/mkp/scsi/c/02c66326dc7e [4/5] scsi_transport_fc: Added store fucntionality to set the rport port_state using sysfs https://git.kernel.org/mkp/scsi/c/afdd11269400 [5/5] scsi:lpfc: Added support for eh_should_retry_cmd https://git.kernel.org/mkp/scsi/c/7f3a79a7fd51 -- Martin K. Petersen Oracle Linux Engineering
This patch adds a support to prevent retries of all the io's after an abort succeeds on a particular device when transport connectivity to the device is encountering intermittent errors. Intermittent connectivity is a condition that can be detected by transport fabric notifications. A service can monitor the ELS notifications and take action on all the outstanding io's of a scsi device at that instant. This feature is intended to be used when the device is part of a multipath environment. When the service detects the poor connectivity, the multipath path can be placed in a marginal path group and ignored further io operations. After placing a path in the marginal path group,the daemon sets the port_state to Marginal which sets bit in scmd->state for all the io's on that particular device with the new sysfs interface provided in this patch.This prevent retries of all the io's if an io hits a scsi timeout which inturn issues an abort. On Abort succeeds on a marginal path the io will be immediately retried on another active path.On abort fails then the things escalates to existing target reset sg interface recovery process. Below is the interface provided to set the port state to Marginal and Online. echo "Marginal" >> /sys/class/fc_remote_ports/rport-X\:Y-Z/port_state echo "Online" >> /sys/class/fc_remote_ports/rport-X\:Y-Z/port_state The patches were cut against 5.11-rc2 --- v8: Rebase the patches on top of 5.11-rc2 v7: Added New routine in scsi_host_template to decide if a cmd is retryable instead of checking the same using SCMD_NORETRIES_ABORT bit as the cmd retry part can be checked by validating the port state. Removed the changes related to SCMD_NORETRIES_ABORT bit. Added a new function fc_eh_should_retry_cmd to check whether the cmd should be retried based on the rport state. Reoreder the patch The patches were cut against 5.11/scsi-queue tree v6: Reordered the patches to make patch ordering and more logical. v5: Added the DID_TRANSPORT_MARGINAL case to scsi_decide_disposition Made changes to clear the SCMD_NORETRIES_ABORT bit if the port_state has changed from marginal to online due to port_delete and port_add as we need the normal cmd retry behaviour while we are calling the eh handlers. Made changes in fc_scsi_scan_rport as we are checking FC_PORTSTATE_ONLINE instead of FC_PORTSTATE_ONLINE and FC_PORTSTATE_MARGINAL v4: Made changes in fc_eh_timed_out callout to set the SCMD_NORETRIES_ABORT if port state is marginal With this change, we removed the code to loop over running commands and fc_remote_port_chkready changes to set the SCMD_NORETRIES_ABORT Removed the scsi_cmd argument for fc_remote_port_chkready and reverted back the patches that addressed this change(argument) Removed unnecessary comments Handle the return of errors on failure. v3: Removed the port_state from starget attributes. Enabled the store functionality for port_state under remote port Added a new argument to scsi_cmd to fc_remote_port_chkready Used the existing scsi command iterators scsi_host_busy_iter. Rearranged the patches Added new patches to add new argument for fc_remote_port_chkready v2: Added new error code DID_TRANSPORT_MARGINAL to handle marginal errors. Added a new rport_state FC_PORTSTATE_MARGINAL and also added a new sysfs interface port_state to set the port_state to marginal. Added the support in lpfc to handle the marginal state. *** BLURB HERE *** Muneendra (5): scsi: Added a new error code DID_TRANSPORT_MARGINAL in scsi.h scsi: No retries on abort success scsi_transport_fc: Added a new rport state FC_PORTSTATE_MARGINAL scsi_transport_fc: Added store fucntionality to set the rport port_state using sysfs scsi:lpfc: Added support for eh_should_retry_cmd drivers/scsi/lpfc/lpfc_scsi.c | 1 + drivers/scsi/scsi_error.c | 23 +++++- drivers/scsi/scsi_lib.c | 1 + drivers/scsi/scsi_transport_fc.c | 118 ++++++++++++++++++++++++++----- include/scsi/scsi.h | 1 + include/scsi/scsi_host.h | 6 ++ include/scsi/scsi_transport_fc.h | 4 +- 7 files changed, 133 insertions(+), 21 deletions(-)