CacheManage | 🔥android缓存管理器，分为内存缓存和文件缓存两种 先取内存数据，没有再从文件缓存中取 | Caching library

Two processes with the (default) cluster name dev have found each other and attempted to cluster together. Versions are incompatible so this fails.

You could turn off discovery on both processes, but if you don't control the other this may not be viable.

Instead you could use config.getGroupConfig().setName('...') to change the name of your cluster to prevent the join attempt.

The issue is your MessageSourceComparator, after deserialization it sorts differently.

There was a change how TreeMap is handled between 3.x and 4.x.

In 3.x the TreeMap is serialized using plain Java serialization. Apparently, it deserializes the data in the order it was stored in the map.

In 4.x+ a special serializer for TreeMap was added, what this serializer does on deserialization is that it creates new TreeSet with deserialized comparator and adds all elements to it. Now because your deserialized comparator is different the elements end up in the wrong order.

I don't think it is reasonable to expect to keep the order when the comparator changes on ser/de. What you can do is to cache LinkedHashMap instead:

You can add multiple conditional @Cacheable annotations to the @Caching annotation.

If you do the math:

4s / 250 000 lookups is 0.016 ms per local lookup. This seems rather high, but let's take that.

When you add a single node then the data gets partitioned and half of the requests will be served from the other node. If you add 2 more nodes (4 total) then 25 % of the requests will be served locally and 75 % will be served over network. This should explain why the response time grows when you add more nodes.

Even simple ping on localhost takes twice or more time. On a real network the read latency we see in benchmarks is 0.3-0.4 ms per read call. This makes:

0.25 * 250k *0.016 + 0.75 * 250k * 0.3 = ~57 s

You simply won't be able to make so many calls serially over the network (even local one), you need to either

• parallelize the calls - use javax.cache.Cache#getAll to reduce the number of calls
• you can try enabling reading local backups via com.hazelcast.config.MapConfig#setReadBackupData so there is less requests over the network.

The read backup data feature is only available for IMap, so you would need to use Spring caching with hazelcast-spring module and its com.hazelcast.spring.cache.HazelcastCacheManager:

With some more research over the web, I was able to fix the same exception which was occurring in my project by adding spring-context-support to the project. I did it by adding following dependency to my build.gradle:

No, it is not

Do not pass NSManagedObject instances between queues. Doing so can result in corruption of the data and termination of the app. When it is necessary to hand off a managed object reference from one queue to another, use NSManagedObjectID instances.

You would want something like:

There are Cache, Spring Cache and Map to consider here.

For Cache, Hazelcast is an implementation provider for the Java caching standard, JSR107. These show as "Cache" on the Management Center, and if you run hazelcastInstance.getDistributedObjects() they'll be of type ICache. It's in the Hazelcast documentation here.

For Map, Hazelcast provides a data structure IMap which is mostly a superset of java.util.Map. These show as "Map" on the Management Center.

Spring also provides caching, and you can set CacheType for JSR107 or directly with Hazelcast, or allow Spring to pick. When Spring uses Hazelcast directly, it will use IMap not ICache for storage.

If you pick JCache or configure Spring to use JCache, then you get standards compliant behaviour. You have caching, and can easily swap caching provider from Hazelcast to something else should you want to.

Map gives you operations such as executeOnKey to update one or more entries in situ. If the entry is a compound object but on a small part is changing, this can be a more efficient way to update it than sending the whole value.

You need to add the CACHE_MANAGER as a provider to your TestingModule:

you have to follow one of the SOLID principles

The interface segregation principle: "Many client-specific interfaces are better than one general-purpose interface."

or