Fixed wrong synchronization of threads in throttling counting #13
Conversation
There was a problem hiding this comment.
From my side, it is necessary to do:
- read contribution notes - https://openhubframework.atlassian.net/wiki/display/OHF/Contribution
- read contribution rules and principles - https://openhubframework.atlassian.net/wiki/display/OHF/Rules+and+principles
- for all java based files please change version to 2.0.0 (next release version will be 2.0.0)
- please avoid to use full author tag (do not add e-mail address)
- it is necessary to rebase these commits on to develop and solve some conflicts
- please improve title and description of this PR
- it is possible to squash all commits together?
After I can continue in review.
There was a problem hiding this comment.
Please add summary of what is going on in this task (https://openhubframework.atlassian.net/browse/OHFJIRA-14) - we know problem and it's necessary to add task description, what is implemented
| package org.openhubframework.openhub.common.synchronization; | ||
|
|
||
| /** | ||
| * Method {@link #doInSynchronizationPart()} in this interface can be synchronized by one one value in |
| import java.util.concurrent.ConcurrentHashMap; | ||
|
|
||
| /** | ||
| * Executor for {@link SynchronizationBlock} which synchronized more then one threads by one value. |
| * Executor for {@link SynchronizationBlock} which synchronized more then one threads by one value. | ||
| * <p> | ||
| * For synchronization by one value call {@link #execute(SynchronizationBlock, String, Object)}, where in method | ||
| * {@link SynchronizationBlock#doInSynchronizationPart()} are code which will be by synchronized by value. |
There was a problem hiding this comment.
which -> that + will be synchronized by value
| * Method will be synchronized by one value calling | ||
| * {@link SynchronizationExecutor#execute(SynchronizationBlock, String, Object)}. | ||
| */ | ||
| protected abstract void doInSynchronizationPartNoResult(); |
There was a problem hiding this comment.
When the class has name SynchronizationNoResultBlock then method could be shorter, for example doInSynchro or synchroBlock ...
| * <p> | ||
| * For synchronization by one value call {@link #execute(SynchronizationBlock, String, Object)}, where in method | ||
| * {@link SynchronizationBlock#doInSynchronizationPart()} are code which will be by synchronized by value. | ||
| * </p> |
|
Maybe we should consider new PR in Camel Core: |
hanusto
changed the title
hanusto
changed the title
| * | ||
| * @return instance of this class | ||
| */ | ||
| public synchronized static SynchronizationExecutor getInstance() { |
There was a problem hiding this comment.
Synchronization of lazy loaded singleton is broken. Please you static eager initialization - http://www.javaworld.com/article/2074979/java-concurrency/double-checked-locking--clever--but-broken.html?page=2.
| @@ -45,81 +43,60 @@ | |||
|
|
|||
| private static final int DUMP_PERIOD = 60; | |||
|
|
|||
| private static final String SYNC_THROTTLING_TYPE = "SYNCHRONIZATION_THROTTLING"; | |||
There was a problem hiding this comment.
Maybe THROTTLING_SYNCHRONIZATION is better name.
|
|
||
| long now = DateTime.now().getMillis(); | ||
| long from = now - (interval * 1000); | ||
| long now = DateTime.now().getMillis(); |
There was a problem hiding this comment.
We use spaces for indent, no tabs - see https://openhubframework.atlassian.net/wiki/display/OHF/Source+code+conventions#Sourcecodeconventions-Syntaxrules.
openhubframework
force-pushed
the
feature/OHFJIRA-14
branch
from
ef48f1c
to
f65747d
Compare
|
Coverage increased (+0.3%) to 70.893% when pulling f65747d774d276a31bace8666e73491127e9d179 on feature/OHFJIRA-14 into 716fd89 on develop. |
Issue: OHFJIRA-14
openhubframework
force-pushed
the
feature/OHFJIRA-14
branch
from
f65747d
to
b30aaf9
Compare
No description provided.