ib: added external send queue to workaround downtime
authorPeter Somogyi <psomogyi@gamax.hu>
Mon, 26 Feb 2007 10:59:20 +0000 (11:59 +0100)
committerPeter Somogyi <psomogyi@gamax.hu>
Mon, 26 Feb 2007 10:59:20 +0000 (11:59 +0100)
commit1714c1ed0ce6508d6cf86aa3bde641f2280d40b4
treee579b9ba0eb80c8d855d4814baebf8bd4b83ba8c
parent6bcb712d3bd9a8a44cb9f70ffcc4fd3e14581fab
ib: added external send queue to workaround downtime
Workaround is because I couldn't find a correct way in ib to reconnect cleanly (with queue kept) when destination is unreachable.
When connection is broken, all internal queue contents are being destroyed and reconnects automatically.
An "external" send queue is kept until the connection is up again for a dest node.

(This used to be ctdb commit 49cac6eafb9d9c80ef1c7f0388aa7786c8be6f24)
ctdb/common/ctdb_message.c
ctdb/ib/ibw_ctdb.c
ctdb/ib/ibw_ctdb.h
ctdb/ib/ibw_ctdb_init.c
ctdb/ib/ibwrapper.c
ctdb/ib/ibwrapper.h
ctdb/ib/ibwrapper_test.c