Ownerless client - client side #15859
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
sancar
force-pushed
the
enh/ownerless_clientSide/master
branch
4 times, most recently
from
2b47680
to
469a87a
Compare
|
run-lab-run |
1 similar comment
|
run-lab-run |
sancar
force-pushed
the
enh/ownerless_clientSide/master
branch
from
a3febc2
to
b8fd5e1
Compare
ihsandemir
requested changes
Nov 5, 2019
...cast/src/main/java/com/hazelcast/client/impl/connection/nio/ClientConnectionManagerImpl.java
Show resolved
Hide resolved
...cast/src/main/java/com/hazelcast/client/impl/connection/nio/ClientConnectionManagerImpl.java
Outdated
Show resolved
Hide resolved
...cast/src/main/java/com/hazelcast/client/impl/connection/nio/ClientConnectionManagerImpl.java
Outdated
Show resolved
Hide resolved
...cast/src/main/java/com/hazelcast/client/impl/connection/nio/ClientConnectionManagerImpl.java
Outdated
Show resolved
Hide resolved
...cast/src/main/java/com/hazelcast/client/impl/connection/nio/ClientConnectionManagerImpl.java
Outdated
Show resolved
Hide resolved
...cast/src/main/java/com/hazelcast/client/impl/connection/nio/ClientConnectionManagerImpl.java
Show resolved
Hide resolved
...cast/src/main/java/com/hazelcast/client/impl/connection/nio/ClientConnectionManagerImpl.java
Outdated
Show resolved
Hide resolved
...cast/src/main/java/com/hazelcast/client/impl/connection/nio/ClientConnectionManagerImpl.java
Outdated
Show resolved
Hide resolved
sancar
force-pushed
the
enh/ownerless_clientSide/master
branch
8 times, most recently
from
c6fe274
to
955bd18
Compare
ihsandemir
approved these changes
Nov 12, 2019
sancar
pushed a commit
to sancar/hazelcast
that referenced
this pull request
Nov 14, 2019
Invocations on cluster restart like registering listeners, creating proxies are all urgent and they are not checked for max invocation count. recreateCachesOnCluster also will not checked with the changes in this pr. fix is backported from hazelcast#15859 fixes hazelcast#15556
|
is this RP a possible fix for #15556 ? |
sancar
pushed a commit
to sancar/hazelcast
that referenced
this pull request
Nov 14, 2019
Invocations on cluster restart like registering listeners, creating proxies are all urgent and they are not checked for max invocation count. recreateCachesOnCluster also will not checked with the changes in this pr. fix is backported from hazelcast#15859 fixes hazelcast#15556
|
@Danny-Hazelcast Yes, this fixes #15556. I have updated the pr description accordingly. |
sancar
pushed a commit
to sancar/hazelcast
that referenced
this pull request
Nov 15, 2019
Invocations on cluster restart like registering listeners, creating proxies are all urgent and they are not checked for max invocation count. recreateCachesOnCluster also will not checked with the changes in this pr. fix is backported from hazelcast#15859 fixes hazelcast#15556
sancar
pushed a commit
to sancar/hazelcast
that referenced
this pull request
Nov 15, 2019
Invocations on cluster restart like registering listeners, creating proxies are all urgent and they are not checked for max invocation count. recreateCachesOnCluster also will not checked with the changes in this pr. fix is backported from hazelcast#15859 fixes hazelcast#15556
sancar
pushed a commit
to sancar/hazelcast
that referenced
this pull request
Nov 15, 2019
Invocations on cluster restart like registering listeners, creating proxies are all urgent and they are not checked for max invocation count. recreateCachesOnCluster also will not checked with the changes in this pr. fix is backported from hazelcast#15859 fixes hazelcast#15556
sancar
force-pushed
the
enh/ownerless_clientSide/master
branch
from
955bd18
to
56a389b
Compare
asimarslan
approved these changes
Nov 21, 2019
sancar
force-pushed
the
enh/ownerless_clientSide/master
branch
3 times, most recently
from
d978c48
to
4bea997
Compare
sancar
force-pushed
the
enh/ownerless_clientSide/master
branch
from
4bea997
to
1d108a1
Compare
sancar
pushed a commit
to sancar/hazelcast
that referenced
this pull request
Dec 4, 2019
Follow up to hazelcast#15859 The client failover behaviour got broken after ownerless client changes. This pr restores the behaviour back. related to https://github.com/hazelcast/hazelcast-enterprise/issues/3385 - Resetting the internal client state explicitly when trying a next cluster. For tests it is too late to wait for a connection to be established and its uuid to be checked. Tests are expecting to see empty memberlist when initial cluster is gone and when a next cluster connection is not established. - We wait for the initial membership events after a cluster change is detected before CLIENT_CONNECTED or CLIENT_CHANGED_CLUSTER event. - Don't fire member removed events and don't print member-list when initial membership has reset. - Only first connection after disconnection should check if a cluster changed or not. Otherwise there was a race and a second connection could prevent the first one to fire Connected/changed events
sancar
pushed a commit
to sancar/hazelcast
that referenced
this pull request
Dec 4, 2019
Follow up to hazelcast#15859 The client failover behaviour got broken after ownerless client changes. This pr restores the behaviour back. related to https://github.com/hazelcast/hazelcast-enterprise/issues/3385 - Resetting the internal client state explicitly when trying a next cluster. For tests it is too late to wait for a connection to be established and its uuid to be checked. Tests are expecting to see empty memberlist when initial cluster is gone and when a next cluster connection is not established. - We wait for the initial membership events after a cluster change is detected before CLIENT_CONNECTED or CLIENT_CHANGED_CLUSTER event. - Don't fire member removed events and don't print member-list when initial membership has reset. - Only first connection after disconnection should check if a cluster changed or not. Otherwise there was a race and a second connection could prevent the first one to fire Connected/changed events
sancar
pushed a commit
to sancar/hazelcast
that referenced
this pull request
Dec 4, 2019
Follow up to hazelcast#15859 The client failover behaviour got broken after ownerless client changes. This pr restores the behaviour back. related to https://github.com/hazelcast/hazelcast-enterprise/issues/3385 - Resetting the internal client state explicitly when trying a next cluster. For tests it is too late to wait for a connection to be established and its uuid to be checked. Tests are expecting to see empty memberlist when initial cluster is gone and when a next cluster connection is not established. - We wait for the initial membership events after a cluster change is detected before CLIENT_CONNECTED or CLIENT_CHANGED_CLUSTER event. - seperate cluster restart and cluster change reset behaviour. When cluster change , we get initial membership event. When cluster restart, we don't get initial membership event. Instead, we get member added and removed events for each action. - Only first connection after disconnection should check if a cluster restarted or not. Otherwise there was a race and a second connection could prevent the first one to fire Connected/changed events
sancar
pushed a commit
to sancar/hazelcast
that referenced
this pull request
Dec 5, 2019
sancar
pushed a commit
to sancar/hazelcast
that referenced
this pull request
Dec 9, 2019
Follow up to hazelcast#15859 The client failover behaviour got broken after ownerless client changes. This pr restores the behaviour back. related to https://github.com/hazelcast/hazelcast-enterprise/issues/3385 - Resetting the internal client state explicitly when trying a next cluster. For tests it is too late to wait for a connection to be established and its uuid to be checked. Tests are expecting to see empty memberlist when initial cluster is gone and when a next cluster connection is not established. - We wait for the initial membership events after a cluster change is detected before CLIENT_CONNECTED or CLIENT_CHANGED_CLUSTER event. - seperate cluster restart and cluster change reset behaviour. When cluster change , we get initial membership event. When cluster restart, we don't get initial membership event. Instead, we get member added and removed events for each action. - Only first connection after disconnection should check if a cluster restarted or not. Otherwise there was a race and a second connection could prevent the first one to fire Connected/changed events
sancar
pushed a commit
to sancar/hazelcast
that referenced
this pull request
Dec 10, 2019
sancar
pushed a commit
to sancar/hazelcast
that referenced
this pull request
Dec 10, 2019
Regression caused by hazelcast#15859 fixes https://github.com/hazelcast/hazelcast-enterprise/issues/3419 (cherry picked from commit 74893d2)
sancar
pushed a commit
that referenced
this pull request
Dec 11, 2019
Regression caused by #15859 fixes https://github.com/hazelcast/hazelcast-enterprise/issues/3419 (cherry picked from commit 74893d2)
sancar
pushed a commit
that referenced
this pull request
Dec 11, 2019
sancar
pushed a commit
that referenced
this pull request
Dec 11, 2019
Follow up to #15859 related to https://github.com/hazelcast/hazelcast-enterprise/issues/3385 - Add cluster listener to single connection as before - Resetting the internal client state explicitly when trying a next cluster. For tests it is too late to wait for a connection to be established and its uuid to be checked. Tests are expecting to see empty memberlist when initial cluster is gone and when a next cluster connection is not established. - We wait for the initial membership events after a cluster change is detected before CLIENT_CONNECTED or CLIENT_CHANGED_CLUSTER event. - seperate cluster restart and cluster change reset behaviour. When cluster change , we get initial membership event. When cluster restart, we don't get initial membership event. Instead, we get member added and removed events for each action. - Only first connection after disconnection should check if a cluster restarted or not. Otherwise there was a race and a second connection could prevent the first one to fire Connected/changed events - Test duration fixes. use setClusterConnectTimeoutMillis instead of setMaxBackoffMillis - Event handlers is moved to ClientConnection to avoid racy situations when tracking lastCorrelation id to clean remove event handler. - ConnectionManager.getSize(MemberSelector) is removed since no usage is left. - ClientClusterViewListener handles a failed registration attempt by trying reregister on a random connection.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Related protocol changes
hazelcast/hazelcast-client-protocol#254
==EDIT==
fixes #12841
fixes #15556