[lustre-devel] [PATCH 4/8] lustre: fld: retry fld rpc even for ESHUTDOWN

[Andreas Dilger]

This is mostly useful only on the servers, but clients also allocate "sequences" of 1, so it doesn't hurt to apply it to clients as well. 

Cheers, Andreas

> On Jul 24, 2019, at 19:44, James Simmons <jsimmons at infradead.org> wrote:
> 
> From: wang di <di.wang at intel.com>
> 
> when LWP is being evicted, because it is not replayable,
> the request might return ESHUTDOWN or EWOULDBLOCK, instead
> of failed, which might cause application failure, fld
> client will retry RPC, until the connection is being setup
> again or the LWP is being closed.
> 
> WC-bug-id: https://jira.whamcloud.com/browse/LU-4420
> Lustre-commit: d335e310d4bf490509998ddbb1824e38cff20998
> Signed-off-by: wang di <di.wang at intel.com>
> Reviewed-on: http://review.whamcloud.com/10285
> Reviewed-by: Fan Yong <fan.yong at intel.com>
> Reviewed-by: Johann Lombardi <johann.lombardi at intel.com>
> ---
> fs/lustre/fld/fld_request.c | 12 ++++++------
> 1 file changed, 6 insertions(+), 6 deletions(-)
> 
> diff --git a/fs/lustre/fld/fld_request.c b/fs/lustre/fld/fld_request.c
> index ec45ea6..ba0ef82 100644
> --- a/fs/lustre/fld/fld_request.c
> +++ b/fs/lustre/fld/fld_request.c
> @@ -367,12 +367,12 @@ int fld_client_rpc(struct obd_export *exp,
>    rc = ptlrpc_queue_wait(req);
>    obd_put_request_slot(&exp->exp_obd->u.cli);
>    if (rc != 0) {
> -        if (rc == -EWOULDBLOCK) {
> -            /* For no_delay req(see above), EWOULDBLOCK means the
> -             * connection is being evicted, but this seq lookup
> -             * should not return error, since it would cause
> -             * unecessary failure of the application, instead
> -             * it should retry here
> +        if (rc == -EWOULDBLOCK || rc == -ESHUTDOWN) {
> +            /* For no_delay req(see above), EWOULDBLOCK and
> +             * ESHUTDOWN means the connection is being evicted,
> +             * but this seq lookup should not return error,
> +             * since it would cause unecessary failure of the
> +             * application, instead it should retry here
>             */
>            ptlrpc_req_finished(req);
>            goto again;
> -- 
> 1.8.3.1
>