This is an automated email from the ASF dual-hosted git repository.
dcapwell pushed a commit to branch trunk
in repository https://gitbox.apache.org/repos/asf/cassandra.git
The following commit(s) were added to refs/heads/trunk by this push:
new 9184dd5a99 When doing a host replacement, we need to check that the
node is a live node before failing with "Cannot replace a live node..."
9184dd5a99 is described below
commit 9184dd5a998366dc2b5c18d4954b13b033efcf80
Author: Francisco Guerrero <[email protected]>
AuthorDate: Mon Aug 15 09:23:56 2022 -0700
When doing a host replacement, we need to check that the node is a live
node before failing with "Cannot replace a live node..."
patch by Francisco Guerrero; reviewed by Brandon Williams, David Capwell
for CASSANDRA-17805
---
CHANGES.txt | 1 +
src/java/org/apache/cassandra/service/StorageService.java | 13 +++++++++++--
2 files changed, 12 insertions(+), 2 deletions(-)
diff --git a/CHANGES.txt b/CHANGES.txt
index dd09b25c56..7dce2b90bd 100644
--- a/CHANGES.txt
+++ b/CHANGES.txt
@@ -1,4 +1,5 @@
4.2
+ * When doing a host replacement, we need to check that the node is a live
node before failing with "Cannot replace a live node..." (CASSANDRA-17805)
* Add support to generate a One-Shot heap dump on unhandled exceptions
(CASSANDRA-17795)
* Rate-limit new client connection auth setup to avoid overwhelming bcrypt
(CASSANDRA-17812)
* DataOutputBuffer#scratchBuffer can use off-heap or on-heap memory as a
means to control memory allocations (CASSANDRA-16471)
diff --git a/src/java/org/apache/cassandra/service/StorageService.java
b/src/java/org/apache/cassandra/service/StorageService.java
index ba04ccec91..b7756c0c21 100644
--- a/src/java/org/apache/cassandra/service/StorageService.java
+++ b/src/java/org/apache/cassandra/service/StorageService.java
@@ -1855,14 +1855,23 @@ public class StorageService extends
NotificationBroadcasterSupport implements IE
}
// check for operator errors...
+ long nanoDelay = MILLISECONDS.toNanos(ringTimeoutMillis);
for (Token token : bootstrapTokens)
{
InetAddressAndPort existing =
tokenMetadata.getEndpoint(token);
if (existing != null)
{
- long nanoDelay = ringTimeoutMillis * 1000000L;
- if
(Gossiper.instance.getEndpointStateForEndpoint(existing).getUpdateTimestamp() >
(nanoTime() - nanoDelay))
+ EndpointState endpointStateForExisting =
Gossiper.instance.getEndpointStateForEndpoint(existing);
+ long updateTimestamp =
endpointStateForExisting.getUpdateTimestamp();
+ long allowedDelay = nanoTime() - nanoDelay;
+
+ // if the node was updated within the ring delay or
the node is alive, we should fail
+ if (updateTimestamp > allowedDelay ||
endpointStateForExisting.isAlive())
+ {
+ logger.error("Unable to replace node for token={}.
The node is reporting as {}alive with updateTimestamp={}, allowedDelay={}",
+ token,
endpointStateForExisting.isAlive() ? "" : "not ", updateTimestamp,
allowedDelay);
throw new UnsupportedOperationException("Cannot
replace a live node... ");
+ }
collisions.add(existing);
}
else
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
