This prevents rpc.statd's in-memory (and on-disk) monitor lists from
winding up with multiple records for the same peer with outdated
cookie values. This happens in some HA-NFS configurations where
rpc.statd is always running.
Signed-off-by: Scott Mayhew <[email protected]>
---
utils/statd/monitor.c | 40 +++++++++++++++++++++++++++-------------
1 file changed, 27 insertions(+), 13 deletions(-)
diff --git a/utils/statd/monitor.c b/utils/statd/monitor.c
index 286a5e2..368bd80 100644
--- a/utils/statd/monitor.c
+++ b/utils/statd/monitor.c
@@ -72,6 +72,7 @@ sm_mon_1_svc(struct mon *argp, struct svc_req *rqstp)
.sin_addr.s_addr = htonl(INADDR_LOOPBACK),
};
char *dnsname = NULL;
+ int existing = 0;
xlog(D_CALL, "Received SM_MON for %s from %s", mon_name, my_name);
@@ -148,17 +149,26 @@ sm_mon_1_svc(struct mon *argp, struct svc_req *rqstp)
if (statd_matchhostname(NL_MY_NAME(clnt), my_name) &&
NL_MY_PROC(clnt) == id->my_proc &&
NL_MY_PROG(clnt) == id->my_prog &&
- NL_MY_VERS(clnt) == id->my_vers &&
- memcmp(NL_PRIV(clnt), argp->priv, SM_PRIV_SIZE) == 0) {
- /* Hey! We already know you guys! */
- xlog(D_GENERAL,
- "Duplicate SM_MON request for %s "
- "from procedure on %s",
- mon_name, my_name);
+ NL_MY_VERS(clnt) == id->my_vers) {
+ if (memcmp(NL_PRIV(clnt), argp->priv, SM_PRIV_SIZE)) {
+ xlog(D_GENERAL,
+ "Received SM_MON request with new "
+ "cookie for %s from procedure on %s",
+ mon_name, my_name);
+
+ existing = 1;
+ break;
+ } else {
+ /* Hey! We already know you guys! */
+ xlog(D_GENERAL,
+ "Duplicate SM_MON request for %s "
+ "from procedure on %s",
+ mon_name, my_name);
- /* But we'll let you pass anyway. */
- free(dnsname);
- goto success;
+ /* But we'll let you pass anyway. */
+ free(dnsname);
+ goto success;
+ }
}
clnt = NL_NEXT(clnt);
}
@@ -167,7 +177,7 @@ sm_mon_1_svc(struct mon *argp, struct svc_req *rqstp)
* We're committed...ignoring errors. Let's hope that a malloc()
* doesn't fail. (I should probably fix this assumption.)
*/
- if (!(clnt = nlist_new(my_name, mon_name, 0))) {
+ if (!existing && !(clnt = nlist_new(my_name, mon_name, 0))) {
free(dnsname);
xlog_warn("out of memory");
goto failure;
@@ -180,8 +190,11 @@ sm_mon_1_svc(struct mon *argp, struct svc_req *rqstp)
clnt->dns_name = dnsname;
/*
- * Now, Create file on stable storage for host.
+ * Now, Create file on stable storage for host, first deleting any
+ * existing records on file.
*/
+ nsm_delete_monitored_host(dnsname, mon_name, my_name);
+
if (!nsm_insert_monitored_host(dnsname,
(struct sockaddr *)(char *)&my_addr, argp)) {
nlist_free(NULL, clnt);
@@ -190,7 +203,8 @@ sm_mon_1_svc(struct mon *argp, struct svc_req *rqstp)
/* PRC: do the HA callout: */
ha_callout("add-client", mon_name, my_name, -1);
- nlist_insert(&rtnl, clnt);
+ if (!existing)
+ nlist_insert(&rtnl, clnt);
xlog(D_GENERAL, "MONITORING %s for %s", mon_name, my_name);
success:
result.res_stat = STAT_SUCC;
--
2.4.3
On 11/20/2015 04:27 PM, Scott Mayhew wrote:
> This prevents rpc.statd's in-memory (and on-disk) monitor lists from
> winding up with multiple records for the same peer with outdated
> cookie values. This happens in some HA-NFS configurations where
> rpc.statd is always running.
>
> Signed-off-by: Scott Mayhew <[email protected]>
Committed...
steved.
> ---
> utils/statd/monitor.c | 40 +++++++++++++++++++++++++++-------------
> 1 file changed, 27 insertions(+), 13 deletions(-)
>
> diff --git a/utils/statd/monitor.c b/utils/statd/monitor.c
> index 286a5e2..368bd80 100644
> --- a/utils/statd/monitor.c
> +++ b/utils/statd/monitor.c
> @@ -72,6 +72,7 @@ sm_mon_1_svc(struct mon *argp, struct svc_req *rqstp)
> .sin_addr.s_addr = htonl(INADDR_LOOPBACK),
> };
> char *dnsname = NULL;
> + int existing = 0;
>
> xlog(D_CALL, "Received SM_MON for %s from %s", mon_name, my_name);
>
> @@ -148,17 +149,26 @@ sm_mon_1_svc(struct mon *argp, struct svc_req *rqstp)
> if (statd_matchhostname(NL_MY_NAME(clnt), my_name) &&
> NL_MY_PROC(clnt) == id->my_proc &&
> NL_MY_PROG(clnt) == id->my_prog &&
> - NL_MY_VERS(clnt) == id->my_vers &&
> - memcmp(NL_PRIV(clnt), argp->priv, SM_PRIV_SIZE) == 0) {
> - /* Hey! We already know you guys! */
> - xlog(D_GENERAL,
> - "Duplicate SM_MON request for %s "
> - "from procedure on %s",
> - mon_name, my_name);
> + NL_MY_VERS(clnt) == id->my_vers) {
> + if (memcmp(NL_PRIV(clnt), argp->priv, SM_PRIV_SIZE)) {
> + xlog(D_GENERAL,
> + "Received SM_MON request with new "
> + "cookie for %s from procedure on %s",
> + mon_name, my_name);
> +
> + existing = 1;
> + break;
> + } else {
> + /* Hey! We already know you guys! */
> + xlog(D_GENERAL,
> + "Duplicate SM_MON request for %s "
> + "from procedure on %s",
> + mon_name, my_name);
>
> - /* But we'll let you pass anyway. */
> - free(dnsname);
> - goto success;
> + /* But we'll let you pass anyway. */
> + free(dnsname);
> + goto success;
> + }
> }
> clnt = NL_NEXT(clnt);
> }
> @@ -167,7 +177,7 @@ sm_mon_1_svc(struct mon *argp, struct svc_req *rqstp)
> * We're committed...ignoring errors. Let's hope that a malloc()
> * doesn't fail. (I should probably fix this assumption.)
> */
> - if (!(clnt = nlist_new(my_name, mon_name, 0))) {
> + if (!existing && !(clnt = nlist_new(my_name, mon_name, 0))) {
> free(dnsname);
> xlog_warn("out of memory");
> goto failure;
> @@ -180,8 +190,11 @@ sm_mon_1_svc(struct mon *argp, struct svc_req *rqstp)
> clnt->dns_name = dnsname;
>
> /*
> - * Now, Create file on stable storage for host.
> + * Now, Create file on stable storage for host, first deleting any
> + * existing records on file.
> */
> + nsm_delete_monitored_host(dnsname, mon_name, my_name);
> +
> if (!nsm_insert_monitored_host(dnsname,
> (struct sockaddr *)(char *)&my_addr, argp)) {
> nlist_free(NULL, clnt);
> @@ -190,7 +203,8 @@ sm_mon_1_svc(struct mon *argp, struct svc_req *rqstp)
>
> /* PRC: do the HA callout: */
> ha_callout("add-client", mon_name, my_name, -1);
> - nlist_insert(&rtnl, clnt);
> + if (!existing)
> + nlist_insert(&rtnl, clnt);
> xlog(D_GENERAL, "MONITORING %s for %s", mon_name, my_name);
> success:
> result.res_stat = STAT_SUCC;
>