|
 |
dc8c34 |
From fc202a17f24ecb585680b4a31586d7eec03489a7 Mon Sep 17 00:00:00 2001
|
|
|
dc8c34 |
From: Thierry Bordaz <tbordaz@redhat.com>
|
|
|
dc8c34 |
Date: Wed, 10 Feb 2016 15:17:02 +0100
|
|
|
dc8c34 |
Subject: [PATCH 378/378] Ticket 48445: keep alive entries can break
|
|
|
dc8c34 |
replication
|
|
|
dc8c34 |
|
|
|
dc8c34 |
Bug Description:
|
|
|
dc8c34 |
On the consumer side, at the end of a total update the replica is enabled and the changelog recreated.
|
|
|
dc8c34 |
When the replica is enabled the keep alive entry (for that replica) is created .
|
|
|
dc8c34 |
There is a race condition (that look quite systematic in our tests) if the creation of the entry is added to the changelog
|
|
|
dc8c34 |
before the changelog is recreated.
|
|
|
dc8c34 |
In that case the ADD is erased from the CL and will never be replicated.
|
|
|
dc8c34 |
|
|
|
dc8c34 |
The keep alive entry is created (if it does not already exist) :
|
|
|
dc8c34 |
- during a total update (as supplier)
|
|
|
dc8c34 |
- when the keep alive is updated
|
|
|
dc8c34 |
- when the replica is enabled
|
|
|
dc8c34 |
|
|
|
dc8c34 |
Fix Description:
|
|
|
dc8c34 |
It is not strictly necessary to create the keep alive when the replica is enabled.
|
|
|
dc8c34 |
So we can skip the creation during that step.
|
|
|
dc8c34 |
|
|
|
dc8c34 |
https://fedorahosted.org/389/ticket/48445
|
|
|
dc8c34 |
|
|
|
dc8c34 |
Reviewed by: Mark Reynolds (thank you Mark)
|
|
|
dc8c34 |
|
|
|
dc8c34 |
Platforms tested: F23
|
|
|
dc8c34 |
|
|
|
dc8c34 |
Flag Day: no
|
|
|
dc8c34 |
|
|
|
dc8c34 |
Doc impact: no
|
|
|
dc8c34 |
|
|
|
dc8c34 |
(cherry picked from commit 71a891f0dcfd1aafeb3913279d42e33ed2355312)
|
|
|
dc8c34 |
(cherry picked from commit 02af085c2a9c23536c8d276ee35794ec6efc81f5)
|
|
|
dc8c34 |
(cherry picked from commit 3bbcf2071788e0ff3194c4803b06deec67054fe7)
|
|
|
dc8c34 |
(cherry picked from commit 70e6c011a481b61830120568e5449ef4dfe48276)
|
|
|
dc8c34 |
---
|
|
|
dc8c34 |
ldap/servers/plugins/replication/repl5_replica.c | 1 -
|
|
|
dc8c34 |
1 file changed, 1 deletion(-)
|
|
|
dc8c34 |
|
|
|
dc8c34 |
diff --git a/ldap/servers/plugins/replication/repl5_replica.c b/ldap/servers/plugins/replication/repl5_replica.c
|
|
|
dc8c34 |
index 1852cf6..c5e38ee 100644
|
|
|
dc8c34 |
--- a/ldap/servers/plugins/replication/repl5_replica.c
|
|
|
dc8c34 |
+++ b/ldap/servers/plugins/replication/repl5_replica.c
|
|
|
dc8c34 |
@@ -3772,7 +3772,6 @@ replica_enable_replication (Replica *r)
|
|
|
dc8c34 |
/* What to do ? */
|
|
|
dc8c34 |
}
|
|
|
dc8c34 |
|
|
|
dc8c34 |
- replica_subentry_check(r->repl_root, replica_get_rid(r));
|
|
|
dc8c34 |
/* Replica came back online, Check if the total update was terminated.
|
|
|
dc8c34 |
If flag is still set, it was not terminated, therefore the data is
|
|
|
dc8c34 |
very likely to be incorrect, and we should not restart Replication threads...
|
|
|
dc8c34 |
--
|
|
|
dc8c34 |
2.4.3
|
|
|
dc8c34 |
|