URL: https://github.com/SSSD/sssd/pull/511
Author: fidencio
Title: #511: Do not shutdown KCM/Secrets responders when activities are
happening ...
Action: opened
PR body:
"""
Firstly, I'd like to make it **explicit** that this PR is **missing tests**,
but I won't write them down till we have an agreement whether the proposed
patches do look right/good.
Basically, while trying to reproduce https://pagure.io/SSSD/sssd/issue/3470
I've noticed that both secrets and kcm responders were going down due to the
responder_idle_timeout expiring ... even with a lot of activities happening on
them.
Does this approach look right? If yes, then, what would be the easiest way to
test:
- A responder actually goes down after x seconds;
- Any activity on that responder will make the responder alive for more x
seconds;
"""
To pull the PR as Git branch:
git remote add ghsssd https://github.com/SSSD/sssd
git fetch ghsssd pull/511/head:pr511
git checkout pr511
From 67a6e7e671d32a001921fed8a54c4416f9a45f81 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Fabiano=20Fid=C3=AAncio?= <fiden...@redhat.com>
Date: Wed, 7 Feb 2018 13:20:31 +0100
Subject: [PATCH 1/2] SECRETS: reset last_request_time on any activity
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
As all the activities are being handled by the secrets responder itself
and not by responder's common code, we have to take care of re-setting
the last_request_time by ourselves here.
Without this patch, the responder would be shot down after reaching the
idle_timeout with activities happening or not.
Signed-off-by: Fabiano Fidêncio <fiden...@redhat.com>
---
src/responder/secrets/secsrv_cmd.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/src/responder/secrets/secsrv_cmd.c b/src/responder/secrets/secsrv_cmd.c
index fa5970504..1b405a23e 100644
--- a/src/responder/secrets/secsrv_cmd.c
+++ b/src/responder/secrets/secsrv_cmd.c
@@ -588,6 +588,9 @@ static void sec_fd_handler(struct tevent_context *ev,
errno_t ret;
struct cli_ctx *cctx = talloc_get_type(ptr, struct cli_ctx);
+ /* Always reset the responder idle timer on any activity */
+ cctx->rctx->last_request_time = time(NULL);
+
/* Always reset the idle timer on any activity */
ret = reset_client_idle_timer(cctx);
if (ret != EOK) {
From 421a7dd7008141fdd9cc1fe1a4c15126f7578281 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Fabiano=20Fid=C3=AAncio?= <fiden...@redhat.com>
Date: Wed, 7 Feb 2018 13:24:05 +0100
Subject: [PATCH 2/2] KCM: reset last_request_time on any activity
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
As all the activities are being handled by the kcm responder itself and
not by responder's common code, we have to take care of re-setting the
last_request_time by ourselves here.
Without this patch, the responder would be shot down after reaching the
idle_timeout with activities happening or not.
Signed-off-by: Fabiano Fidêncio <fiden...@redhat.com>
---
src/responder/kcm/kcmsrv_cmd.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/src/responder/kcm/kcmsrv_cmd.c b/src/responder/kcm/kcmsrv_cmd.c
index 0b933f0b4..903c89417 100644
--- a/src/responder/kcm/kcmsrv_cmd.c
+++ b/src/responder/kcm/kcmsrv_cmd.c
@@ -588,6 +588,9 @@ static void kcm_fd_handler(struct tevent_context *ev,
errno_t ret;
struct cli_ctx *cctx = talloc_get_type(ptr, struct cli_ctx);
+ /* Always reset the responder idle timer on any activity */
+ cctx->rctx->last_request_time = time(NULL);
+
/* Always reset the idle timer on any activity */
ret = reset_client_idle_timer(cctx);
if (ret != EOK) {
_______________________________________________
sssd-devel mailing list -- sssd-devel@lists.fedorahosted.org
To unsubscribe send an email to sssd-devel-le...@lists.fedorahosted.org
