Ocfs2 devel - Sep 2017 - [patch]o2net: fix qs_holds may could not be zero

Hi all
In our test, We fond that , when the network down, qs->qs_holds may could not
be reduce to zero, lead to the node can't do fence.
o2net_idle_timer -> o2quo_conn_err -> qs->qs_holds++, after
O2NET_QUORUM_DELAY_MS if qs_holds could be subtract to zero, it could do
make_decision.
But if there are many nodes, when one node network down which contains o2net
connections may not do o2net_idle_timer at the same time, so when a o2net_node
have done nn->nn_still_up, but the qs_holds is not be zero, because other
o2net_node have not done nn->nn_still_up.so the first o2net_node will do
o2net_idle_timer again, and the qs_holds could be add again. And the qs_holds is
global variable, so it formed a loop, the node could not do o2quo_make_decision,
because of qs_holds never be zero.

I alter two functions o2quo_conn_up and o2quo_conn_err, take o2quo_set_hold or
o2quo_clear_hold under control of the bit map qs_conn_bm


Best Regards,
Yang Zhang


@@ -280,20 +280,24 @@ void o2quo_conn_up(u8 node)
       struct o2quo_state *qs = &o2quo_state;
        spin_lock(&qs->qs_lock);
-
-        qs->qs_connected++;
-        mlog_bug_on_msg(qs->qs_connected == O2NM_MAX_NODES,
-                         "node %u\n", node);
-        mlog_bug_on_msg(test_bit(node, qs->qs_conn_bm), "node
%u\n", node);
-        set_bit(node, qs->qs_conn_bm);
-
-        mlog(0, "node %u, %d total\n", node, qs->qs_connected);
-
-        if (!test_bit(node, qs->qs_hb_bm))
-                 o2quo_set_hold(qs, node);
-        else
-                 o2quo_clear_hold(qs, node);
-
+       if (!test_bit(node, qs->qs_conn_bm)) {
+                qs->qs_connected++;
+                mlog_bug_on_msg(qs->qs_connected == O2NM_MAX_NODES,
+                            "node %u\n", node);
+                mlog_bug_on_msg(test_bit(node, qs->qs_conn_bm), "node
%u\n", node);
+                set_bit(node, qs->qs_conn_bm);
+
+                mlog(0, "node %u, %d total\n", node,
qs->qs_connected);
+
+                if (!test_bit(node, qs->qs_hb_bm))
+                          o2quo_set_hold(qs, node);
+                else
+                          o2quo_clear_hold(qs, node);
+       }
       spin_unlock(&qs->qs_lock);
}
@@ -314,13 +318,13 @@ void o2quo_conn_err(u8 node)
                                   node, qs->qs_connected);
                 clear_bit(node, qs->qs_conn_bm);
+
+                if (test_bit(node, qs->qs_hb_bm))
+                          o2quo_set_hold(qs, node);
       }
        mlog(0, "node %u, %d total\n", node, qs->qs_connected);
-        if (test_bit(node, qs->qs_hb_bm))
-                 o2quo_set_hold(qs, node);
-
       spin_unlock(&qs->qs_lock);
}
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://oss.oracle.com/pipermail/ocfs2-devel/attachments/20170911/9114b403/attachment.html

Hello Yang,

Thanks for code submit.
But, could you follow the Linux kernel patch format rule?
How to submit a kernel patch, you can refer the doc
https://urldefense.proofpoint.com/v2/url?u=https-3A__www.ozlabs.org_-7Eakpm_stuff_tpp.txt&d=DwIFAg&c=RoP1YumCXCgaWHvlZYR8PQcxBKCX5YTpkKY057SbK10&r=f4ohdmGrYxZejY77yzx3eNgTHb1ZAfZytktjHqNVzc8&m=3nQJc2n3niuZrCBcfTafrMHONl3nSUmUNmSa8O4bs04&s=waaUblcD-UHR0RIXt1Gko92JBn22fbThj1Q0jH65Gy8&e=

Thanks
Gang

>>> 
> Hi all
> In our test, We fond that , when the network down, qs->qs_holds may
could not
> be reduce to zero, lead to the node can't do fence.
> o2net_idle_timer -> o2quo_conn_err -> qs->qs_holds++, after
O2NET_QUORUM_DELAY_MS
> if qs_holds could be subtract to zero, it could do make_decision.
> But if there are many nodes, when one node network down which contains
o2net
> connections may not do o2net_idle_timer at the same time, so when a 
> o2net_node have done nn->nn_still_up, but the qs_holds is not be zero,
because
> other o2net_node have not done nn->nn_still_up.so the first o2net_node
will do
> o2net_idle_timer again, and the qs_holds could be add again. And the
qs_holds
> is global variable, so it formed a loop, the node could not do 
> o2quo_make_decision, because of qs_holds never be zero.
> 
> I alter two functions o2quo_conn_up and o2quo_conn_err, take o2quo_set_hold
> or o2quo_clear_hold under control of the bit map qs_conn_bm
> 
> 
> Best Regards,
> Yang Zhang
> 
> 
> @@ -280,20 +280,24 @@ void o2quo_conn_up(u8 node)
>        struct o2quo_state *qs = &o2quo_state;
>         spin_lock(&qs->qs_lock);
> -
> -        qs->qs_connected++;
> -        mlog_bug_on_msg(qs->qs_connected == O2NM_MAX_NODES,
> -                         "node %u\n", node);
> -        mlog_bug_on_msg(test_bit(node, qs->qs_conn_bm), "node
%u\n", node);
> -        set_bit(node, qs->qs_conn_bm);
> -
> -        mlog(0, "node %u, %d total\n", node,
qs->qs_connected);
> -
> -        if (!test_bit(node, qs->qs_hb_bm))
> -                 o2quo_set_hold(qs, node);
> -        else
> -                 o2quo_clear_hold(qs, node);
> -
> +       if (!test_bit(node, qs->qs_conn_bm)) {
> +                qs->qs_connected++;
> +                mlog_bug_on_msg(qs->qs_connected == O2NM_MAX_NODES,
> +                            "node %u\n", node);
> +                mlog_bug_on_msg(test_bit(node, qs->qs_conn_bm),
"node %u\n",
> node);
> +                set_bit(node, qs->qs_conn_bm);
> +
> +                mlog(0, "node %u, %d total\n", node,
qs->qs_connected);
> +
> +                if (!test_bit(node, qs->qs_hb_bm))
> +                          o2quo_set_hold(qs, node);
> +                else
> +                          o2quo_clear_hold(qs, node);
> +       }
>        spin_unlock(&qs->qs_lock);
> }
> @@ -314,13 +318,13 @@ void o2quo_conn_err(u8 node)
>                                    node, qs->qs_connected);
>                  clear_bit(node, qs->qs_conn_bm);
> +
> +                if (test_bit(node, qs->qs_hb_bm))
> +                          o2quo_set_hold(qs, node);
>        }
>         mlog(0, "node %u, %d total\n", node,
qs->qs_connected);
> -        if (test_bit(node, qs->qs_hb_bm))
> -                 o2quo_set_hold(qs, node);
> -
>        spin_unlock(&qs->qs_lock);
> }