Write-Behind Caching and Re-entrant Calls

Support Team -
     The Coherence User Guide states that:
     "The CacheStore implementation must not call back into the hosting cache service. This includes OR/M solutions that may internally reference Coherence cache services. Note that calling into another cache service instance is allowed, though care should be taken to avoid deeply nested calls (as each call will "consume" a cache service thread and could result in deadlock if a cache service threadpool is exhausted)."
     I have Load-tested a use case wherein I have two caches: ABCache and BACache. ABCache is accessed by the application for write operation, BACache is accessed by the application for read operation. ABCache is a write-behind cache whose CacheStore populates BACache by reversing key and value of each cache entry stored in the ABCache.
     The solution worked under load with no issues.
     But can I use it? Or is it too dangerous?
     My write-behind thread-count setting is left at default (0). The documentation states that
     "If zero, all relevant tasks are performed on the service thread."
     What does this mean? Can I re-enter the caching service if my thread-count is zero?
     Thank you,
     Denis.

Dimitri -
     I am not sure I fully understand your answer:
     1. "Your test worked because write-behing backing map invokes CacheStore methods asynchronously, on a write-behind thread." In my configuration, I have default value for thread-count, which is zero. According to the documentation, that means that CacheStore methods would be executed by the service thread and not by the write-behind thread. Do I understand this correctly?
     2. "If will fail if CacheStore method will need to be invoked synchronously on a service thread." I am not sure what is the purpose of the "service thread". In which scenarios the "CacheStore method will need to be invoked synchronously on a service thread"?
     Thank you,
     Denis.

Similar Messages

  • Write-Behind Caching and Limited Internal Cache Size

    Let's say I have a write-behind cache and configure its internal cache to be of a fixed limited size, e.g. 10000 units. What would happen if more than 10000 units are added to the write-behind cache within the write-delay period? Would my CacheStore's storeAll() get all of the added values or would some of the values be missed because of the internal cache size limitation?

    Hi Denis,     >
         > If an entry is removed while it is still in the
         > write-behind queue, it will be removed from the queue
         > and CacheStore.store(oKey, oValue) will be invoked
         > immediately.
         >
         > Regards,
         > Dimitri
         Dimitri,
         Just to confirm, that I understand it right if there is a queued update to a key which is then remove()-ed from the cache, then the following happens:
         First CacheStore.store(key, queuedUpdateValue) is invoked.
         Afterwards CacheStore.erase(key) is invoked.
         Both synchronously to the remove() call.
         I expected only erase will be invoked.
         BR,
         Robert

  • Write-Behind Caching and Old Values

    Is there a way to access the old value cached in the write-behind cache for the same key from the CacheStore's store() or storeAll() method?

    I have a business POJO with three parts: partA,     > partB, partC inside. Each of these three parts is
         > persisted by a separate SQL. So, every time I persist
         > my POJO, up to 3 SQLs may be executed.
         I understand.
         > When a change happens in my POJO, it goes onto the
         > write-behind queue. In my CacheStore.store() or
         > CacheStore.storeAll() I would like to be able to make
         > an intelligent decision about which of the three
         > parts: partA, partB or partC has actually changed and
         > only run the SQL updates for the changed parts. This
         > would allow me to avoid massive amounts of
         > unnecessary SQL updates for the parts that did not
         > change.
         Right. Keep in mind that there are two conditions that you must be aware of:
         1) Multiple updates could have occurred to the object, meaning that the database update would have to "roll up" the results of multiple changes to the object.
         2) Some or all of the updates could have already occurred to the database. This may be a little trickier to understand, but it reflects the possible machine failure conditions that occurred while a write-behind was in progress.
         Although the latter are unlikely, they should be accounted for, and of course they are harder to test for with certainty. As a result, the updates to the information (the CacheStore implementation) must be built in an "idempotent" manner, i.e. allowing it to be executed more than once with no additional side-effects.
         > If I had access to the POJO stored under the same key
         > before the new value was put in cache, I could use
         > equals() on each of the three parts to find out
         > exactly which one of them changed.
         While this is true, you would need to compare the "known previous database state" version, not just the "old" version.
         > Of course, if this functionality is not available, I
         > would have to create dirty flags for each of the
         > three POJO parts. But I can't really clear my POJO's
         > flags and recache the POJO from within the store() or
         > storeAll(), right?
         Yes, but remember that those flags are "could be dirty" flags, because of the above failure modes that I described.
         Peace,
         Cameron Purdy
         Tangosol Coherence: The Java Data Grid

  • Write-Behind Caching and Multiple Puts

    What happens when two consecutive puts are performed on the write-behind cache for the same key? Will CacheStore's store() or storeAll() be invoked once for every put() or only once for the last put() (the one which overrode the previous cached values)?

    Hi Denis,
         If you use write-behind, there will be no unnesessary database updates - only the last put() will result in database update.
         Regards,
         Dimitri

  • Write behind cache, DB down, when should the system stop taking new data in

    Hello:
    We are trying to use Coherence for our custom ESB, which is brokering payloads of various size between consumer and provider applications.
    Before Coherence, stopping our DB meant organization-wide outage for critically important business services.
    Since we have at least 40G of RAM in production environment, we believe that our app
    can use Coherence write-behind option for tolerating at least several hours worth of DB outage.
    We are currently using a near cache backed by distributed cache in write-behind mode.
    9 business service JVMs (storage enabled=false) use 30 storage enabled JVMs.
    IMPORTANT: We need to create an automated alerting facility determining when
    amount of unsaved data reaches critical level since DB goes down. This alert should help us decide when our application stops accepting inbound traffic.
    It is hard to use QueueSize parameter for that because our payload memory footprint can vary from 1KB to 3MB.
    We do not expire any entries in order to enable support queries against the cache during DB outage.
    Our experiments with trying various flavors of overflow-scheme resulted in OutOfMemoryError, therefore
    we decided to implement RAM-only cache as a first step.
    <near-scheme>
    <scheme-name>message_payload_scheme</scheme-name>
    <front-scheme>
    <local-scheme>
    <scheme-ref>limited_entities_front_scheme</scheme-ref>
    <high-units>100</high-units>
    </local-scheme>
    </front-scheme>
    <back-scheme>
    <distributed-scheme>
    <backing-map-scheme>
    <read-write-backing-map-scheme>
    <internal-cache-scheme>
    <local-scheme>
    <scheme-ref>limited_bytes_scheme</scheme-ref>
    <high-units>199229440</high-units>
    </local-scheme>
    </internal-cache-scheme>
    <cachestore-scheme>
    <class-scheme>
    <class-name>com.comp.MessagePayloadStore</class-name>
    </class-scheme>
    </cachestore-scheme>
    <read-only>false</read-only>
    <write-delay-seconds>3</write-delay-seconds>
    <write-requeue-threshold>2147483646</write-requeue-threshold>
    </read-write-backing-map-scheme>
    </backing-map-scheme>
    <autostart>true</autostart>
    </distributed-scheme>
    </back-scheme>
    </near-scheme>
    <local-scheme>
    <scheme-name>limited_entities_front_scheme</scheme-name>
    <eviction-policy>LRU</eviction-policy>
    <unit-calculator>FIXED</unit-calculator>
    </local-scheme>
    <local-scheme>
    <scheme-name>limited_bytes_scheme</scheme-name>
    <eviction-policy>HYBRID</eviction-policy>
    <unit-calculator>BINARY</unit-calculator>
    </local-scheme>

    Good info ... I feel like I need to restate my original question along with a couple of new questions caused by the discussion above.
    Q1. Does Coherence evict 'dirty', or 'queued', or 'unsaved' objects for cache configuration provided above?
    The answer should be 'NO', otherwise Coherence is unsafe to use as a system of record,
    it should not just drop unsaved information on the floor.
    Q2. What happens to the front tier of the near+partitioned write behind cache described above when amount of unsaved data exceeds max cache capacity defined via high-units?
    I would expect that map.put starts throwing exceptions: cache storage is full, so it should not accept more data
    Q3. How can I determine a moment when amount of dirty data in bytes(!), not in objects, hits 85% of
    max allowed cache capasity configured in bytes (using high-units param and BINARY calculator).
    'DirtyUnits' counter can probably be built with some lower-level Coherence API. Can we use
    this API?
    Please, understand, that we purchased Coherence for reliability, for making our
    system independent from short DB outages, for keeping our business services up
    and running when DBA need some time for admin operations like rebuilding an index.
    Performance benefits are secondary and are not as obvious for our system which
    uses primary keys only and has a well-tuned co-located Oracle back-end.
    We simply cannot put Coherence to production unless we prove that Coherence
    can reliably hold the data and give us information about approaching crisis
    (the cache full of unsaved data).
    If possible, forward this message to Cameron Purdy,
    who was presenting Coherence to our team several moths ago.
    Thanks,
    Vasili Smaliak
    Applications Architect, Enterprise App Integration
    GMAC ResCap
    [email protected]

  • Write-behind cache not removing entries after upgrade to 3.2

    We recently upgraded tangosol.jar and coherence.jar from version 3.0 to version 3.2. After the upgrade, our write-behind caches began consuming all available memory and crashing the JVMs because the entries were not being removed from the cache after being written to the database. We rolled back to the 3.0 jars without making any other modifications and the caches behave as expected. We'd really like to move to 3.2 for the improved network fault tolerance, but we need to resolve this issue first.
    What changes were made in 3.2 with respect to write-behind caches that might cause this issue? I've reviewed our configuration and our code and can't find anything unusual, but I'm not sure what I should be looking for.
    Any ideas?

    I've opened an SR, but I haven't heard back. In the meantime, I've continued digging and I've noticed something strange - in the store() method of our backing map implementation, we take the entry that we just persisted and remove it from the backing map.
    In my small-scale local tests, the size of the map is 1 when we enter store() and is 0 when we leave, as expected. If we process another entry using the 3.0 jars, it's again 1 and then 0. However, it gets more interesting with the 3.2 jars - the size of the map is 1 when we enter store() the first time and 0 when we leave, but if we process another entry, the size is 2 when we enter and 1 when we leave. This pattern continues such that both values increase by 1 every time we process an entry.
    This would imply that we're either removing the entries incorrectly, or they're somehow being reinserted into the map.
    Any ideas?
    Here's the body of our method (with a bunch of sysouts added to the normal logging because this app won't run correctly under a debugger):
            * Store the specified value under the specific key in the underlying
            * store, then remove the specific key from the internal map and hence
            * the cache itself. This method is intended to support both key/value
            * creation and value update for a specific key.
            * @param oKey   key to store the value under
            * @param oValue value to be stored
            * @throws UnsupportedOperationException if this implementation or the
            *                                       underlying store is read-only
            public void store(Object oKey, Object oValue)
                RemoveOnStoreRWBackingMap mapBacking = RemoveOnStoreRWBackingMap.this;
                System.out.println("map storing  " + oKey);
                System.out.println("Size before = " + mapBacking.entrySet().size());
                Iterator entries = mapBacking.entrySet().iterator();
                while (entries.hasNext()) {
                    System.out.println("entry = " + entries.next());   
                String storeClassName = getCacheStore().getClass().getName();
                Logger log = Logger.getLogger(storeClassName);
                log.debug(storeClassName + ": In store method.  Storing " + oKey);
                long cFailuresBefore = getStoreFailures();
                log.debug(storeClassName + ": failures before=" + cFailuresBefore);
                super.store(oKey, oValue);
                long cFailuresAfter = getStoreFailures();
                log.debug(storeClassName + ": failures afer=" + cFailuresAfter);
                if (cFailuresBefore == cFailuresAfter)  {
                    log.debug(storeClassName + ": About to remove");
                    mapBacking = RemoveOnStoreRWBackingMap.this;
                    Converter converter = mapBacking.getContext().getKeyToInternalConverter();
                    System.out.println("removed " + mapBacking.remove(converter.convert(oKey)));
    //                System.out.println("removed " + mapBacking.getInternalCache().remove(converter.convert(oKey)));
                    log.debug(storeClassName + ": Removed");
                Converter converter = RemoveOnStoreRWBackingMap.this.getContext().getKeyFromInternalConverter();
                System.out.println("Size after = " + mapBacking.entrySet().size());
            }

  • Write-Behind, Expiration, and SQL Exceptions.

    Hi Chaps,
    If a cache with write-behind enabled has problems writing to the DB I understand that Coherence will re-queue the objects and write them when the DB is available.
    The problem I have is that (after a DB failure) I don't see them being written - I can see these items in the cache but not in the DB, even several hours after the outage. (Items that were added to the cache after the outage are being written).
    Is there anything the cachestore methods (specifically store() ) need to do with regards to exceptions to ensure that these items are re-qeueued?
    Next question is: I was also wondering how is this managed with regards to expiry?
    We have our own expiry routine which removes items from the cache that are older than 24 hours (this was from before we could expire objects by specifying the timeout in the put() method call, which I am intending to switch to).
    If an item has not been written to the DB due to an outage and is then expired (by our own routine or by Coherence) is it then lost forever, or will it remain in the queue? (seeing as the queue holds references I am guessing not but though I'd check).
    Thanks,
    Randal.

    Jon,
    I have a question related to this...If you remember a few weeks back, I stumbled upon the problem of the "version-persistent" map for the versioned-backing-map-scheme does not accept putAll operations. The workaround until you guys implement it, was to override the putAll method of the cacheStore and throw and unsupported operation exception (to force individual puts).
    Well, although this workaround works, I am getting tons and tons of:
    2006-04-06 17:18:27.347 Tangosol Coherence 3.1/339 <Warning> (thread=WriteBehindThread:MyCacheStore, member=1): The CacheStore "MyCacheStore@46b9979b" does not support storeAll().
    2006-04-06 17:18:27.348 Tangosol Coherence 3.1/339 <Error> (thread=WriteBehindThread:MyCacheStore, member=1): Failed to store keys="[16, 18, 21, 26, 5, 13, 14, 25, 17, 15, 23, 19, 2, 6, 9, 7]":
    java.lang.UnsupportedOperationException
    at ...MyCacheStore.storeAll(MyCacheStore.java:126)
    at com.tangosol.net.cache.ReadWriteBackingMap$CacheStoreWrapper.storeAll(ReadWriteBackingMap.java:3820)
    at com.tangosol.net.cache.ReadWriteBackingMap$WriteThread.run(ReadWriteBackingMap.java:3538)
    at com.tangosol.util.Daemon$1.run(Daemon.java:63)
    2006-04-06 17:18:27.349 Tangosol Coherence 3.1/339 <Warning> (thread=WriteBehindThread:MyCacheStore, member=1): Requeued store for key="16"
    2006-04-06 17:18:27.349 Tangosol Coherence 3.1/339 <Warning> (thread=WriteBehindThread:MyCacheStore, member=1): Requeued store for key="18"
    2006-04-06 17:18:27.350 Tangosol Coherence 3.1/339 <Warning> (thread=WriteBehindThread:MyCacheStore, member=1): Requeued store for key="21"
    2006-04-06 17:18:27.351 Tangosol Coherence 3.1/339 <Warning> (thread=WriteBehindThread:MyCacheStore, member=1): Requeued store for key="26"
    the first OperationNotSupported is expected, but I'm not sure what the requeued warnings are all about. These are not failures to the DB...it is something else. (mind you that this happens when trying to load a lot of data into the map.)
    1- Is this requeuing related or the same as in failed DB stores?
    2- Is it possible to "lose" stores if I don't configure the write-requeue-threshold with very, very high values? I must ensure I don't lose anything.
    In a related note, in some circumstances, I need to ensure that the "write queue" is flushed or cleared. For example, I may want to force a flush of all pending stores (and wait/block until that's done).
    I have looked into it and I don't seem to know how to do it. I can read the write-queue length, but I believe that this is not very accurate...since my tests seem to indicate that the write-behind thread may take the entries to store off the write-queue and then deal with them in parallel (which means that there are still entries althought the write-queue size is 0). Also, there are some calls from the cache store that, at first, seem to give some access to the write thread (potentially allowing me to contact the thread to tell him to flush or discard any pending stores)...but I believe that all of the functions are protected...but there may be other ways..
    I guess my second batch of questions are:
    1- How can I effectively force a flush (or clear) of the pending stores. Such that there is no single store pending in any queue (visible or invisible to the programmer).
    2- What is the role of re-queuing in these situations? where is the queue sitting, the thread? the cache store? who's responsible of retrying that, and when?...I would like to flush those entries too.
    A quick explanation of the operation of the write thread would also be very appreciated.
    Thanks!
    Josep M.

  • Thread pool configuration for write-behind cache store operation?

    Hi,
    Does Coherence have a thread pool configuration for the Coherence CacheStore operation?
    Or the CacheStore implementation needs to do that?
    We're using write-behind and want to use multiple threads to speed up the store operation (storeAll()...)
    Thanks in advance for your help.

    user621063 wrote:
    Hi,
    Does Coherence have a thread pool configuration for the Coherence CacheStore operation?
    Or the CacheStore implementation needs to do that?
    We're using write-behind and want to use multiple threads to speed up the store operation (storeAll()...)
    Thanks in advance for your help.Hi,
    read/write-through operations are carried out on the worker thread (so if you configured a thread-pool for the service the same thread-pool will be used for the cache-store operation).
    for write-behind/read-ahead operations, there is a single dedicated thread per cache above whatever thread-pool is configured, except for remove operations which are synchronous and still carried out on the worker thread (see above).
    All above is of course per storage node.
    Best regards,
    Robert

  • Write behind exception and recovery

    Hi all,
    I am working on write behind part in equity trading system. I know that cache store operation will eventually be thrown away if no of retry exceed write-requeue-threshold. However, this is not acceptable as DB must sync with caches at least at day end. For some more complicated caches, we use cache store implementation and Hiberate for simple cache. I am thinking to capture the sql statements that failed during the day and finally at day end, manually fix issues (egDB issue or others) then have them executed.
    Questions:
    1. Is this a good approach for handling the scenario? If yes, any way I can capture the statements and write to file for running in SQL plus for example in case of Hiberate?
    2. Is there any out of box mechanism in Coherence for recovering write-behind queues in case of WHOLE cluster fail (not node fail).
    Henry

    922963 wrote:
    Hi all,
    I am working on write behind part in equity trading system. I know that cache store operation will eventually be thrown away if no of retry exceed write-requeue-threshold. However, this is not acceptable as DB must sync with caches at least at day end. For some more complicated caches, we use cache store implementation and Hiberate for simple cache. I am thinking to capture the sql statements that failed during the day and finally at day end, manually fix issues (egDB issue or others) then have them executed.
    Questions:
    1. Is this a good approach for handling the scenario? If yes, any way I can capture the statements and write to file for running in SQL plus for example in case of Hiberate?Hi Henry,
    There are a few caveats you need to care about but in general it is possible.
    Not necessarily SQLs but serialized entries would probably be simpler to work with when you try to restore them.
    Also, you have to be aware that Coherence may fail to write an entry to the DB but at retry it may try to write a new entry. If it succeeds, you have to be able to figure that out that the earlier failure must not be reexecuted.
    In effect, you should have per-entry versioning in the database and you should check versions of the entity in the database upon writing both from the cache store and also from your end-of-day retry logic.
    2. Is there any out of box mechanism in Coherence for recovering write-behind queues in case of WHOLE cluster fail (not node fail).
    No, nothing like that comes out-of-the-box, if you lost a partition, you lost your write-behind-enqueued entries, too. You could log your failed writes to disk though as you indicated above.
    Best regards,
    Robert

  • Transactional Caches and Write Through

    I've been trying to implement the use of multiple caches, each with write through, all within a transaction.
         The CacheFactory.commitTransactionCollection(..) method only seems to work correctly if the first transactionMap throws an exception in the database code.
         If the second transactionMap throws exceptions, the caches do not appear to rollback correctly.
         I can wrap the whole operation in a JDBC transaction that rolls back the database correctly but the caches are not all rolled back because they are committed one by one?
         For example, I write to two transaction maps, each one created from separate caches. When commiting the transaction maps, the second transaction map causes a database exception. It appears the first transaction map has already committed its objects and doesn't roll back.
         Is it possible to use Coherence with multiple transaction maps and get all the caches and databases rolled back?
         I've also been trying to look at using coherence-tx.rar as described in the forums within WebLogic but I'm getting @@@@@ Failed to commit: javax.transaction.SystemException: Could not contact coordinator at null+SMARTPC:7001+null+t3+
         (SMARTPC being my pc name)
         Has anybody else had this problem? Bonus points for describing how to fix it!
         Mike

    The transaction support in Coherence is for Local     > Transactions. Basically, what this means is that the
         > first phase of the commit ("prepare") acquires locks
         > and ensures that there are no conflicts. The second
         > phase ("commit") does nothing but push data out to
         > the caches.
         This means that once prepare succeeds (all locks acquired), commit will try to copy local data into the base map. If there is a failure on any put, rollback will undo any changes made. All locks are cleared at the end.
         > The problem is that when you are using a
         > CacheStore module, the exception is occurring during
         > the second phase.
         If you start using a CacheStore module, then database update has to be part of the atomic procedure.
         >
         > For this reason, write-through and cache transactions
         > are not a supported combination.
         This is not true for a cache transaction that updaets a single cache entry, right?
         >
         > For single-cache-entry updates, CacheStore operations
         > are fully fault-tolerant in that the cache and
         > database are guaranteed to be consistent during any
         > server failure (including failures during partial
         > updates). While the mechanisms for fault-tolerance
         > vary, this is true for both write-through and
         > write-behind caches.
         For the write-thru case, I believe Database and cache are atomically updated.
         > Coherence does not support two-phase CacheStore
         > operations across multiple CacheStore instances. In
         > other words, if two cache entries are updated,
         > triggering calls to CacheStore modules sitting on
         > separate servers, it is possible for one database
         > update to succeed and for the other to fail.
         But once we have multiple CacheStore modules, then once one atomic write-thru put succeeds that means database is already updated for that specific put. There is no way to roll back the database update (although we can roll back the cache update). Therefore, you may end up in partial commits in such situations where multiple cache entries are updated across different CacheStore modules.
         If I use write-behind CacheStore modules, I can roll back entirely and avoid partial commits? Since writes are not immediately propagated to the database? So in essence, write-behind cache stores are no different than local transactions... Is my understanding correct?

  • Cache write-behind complete check

    Is there a surefire way to check a cache that has a store persisting objects to the database and write-behind set to 2 seconds, has persisted all objects put into the cache?
    We have tried using JMX, querying the Cache's QueueSize and waiting until it reaches 0. It turns out that when putting objects into the write-behind cache, the write-behind queue is not necessarily non-zero immediately after the put(s). e.g. QueueSize may be 0, even if objects still need to be persisted.
    For our nightly integration tests we need to clear out the cache, but want to make sure we do not call NamedCache.clear() on a cache that still has objects that need to be persisted.
    Any ideas?

    Hi Rob,
    The problem may actually be the timeliness of updates to the QueueSize JMX attribute as we're using the MBeanConnector to obtain information on our cache members (and providers). Assuming that objects actually make it to the write-behind queue during the cache put call and certain tests need to be sure these objects are persisted, instead of doing the accounting approach discussed previously, I found a forum thread on ReadWriteBackingMap flush calls.
    To get access to the ReadWriteBackingMap.flush() I created a small test today using a subclass of ReadWriteBackingMap that registers the backingmap with our CashStore implementation:
        protected void configureCacheStore(CacheStore store, boolean readOnly) {
            super.configureCacheStore(store, readOnly);
            if (store instanceof CacheLoaderWriterProvider) {
                ((CacheLoaderWriterProvider)store).registerBackingMap(this);
        }Our cachstore (CacheLoaderWriterProvider) in turn exposes a call to the registered map's flush method as a JMX operation.
    Whenever we need to be sure the write-behind queue is empty during our tests, we'll call this JMX operation.
    Best Regards,
    Marcel.

  • Can a db slowdown with write-behind cause a slowdown in cache operations?

    If we have a coherence cluster, and one cache configured with write-behind is having trouble writing to the db (ie, it's slow), and we keep adding objects to the cache that exceed the ability of the db to consume them; will flow-control kick in and cause the writes to the cache to block/slow-down? Ie, the classic producer-consumer problem, where we are adding objects to the cache, faster than the cachestore can consume them.
    What happens in this case? Will flow-control kick in and block writes to the cache? Will an internal buffer just keep growing? Are there any knobs to tweak this behavior (eg, in the case of spikes, where temporarily the producer is producing faster than the consumer can consume for a brief period of time, but then things go back to normal)?

    user9222505 wrote:
    I believe we discovered that the same thread pool is used for all requests to the cache, including gets, puts and calls into the cachestore. So if the writes are slow within the cachestore, then it uses up all of the threads and slows everything down.Hi,
    This is not really correct.
    If a cache in a service is configured to use write-behind then a separate thread for that service is started, which deals with write-behind store and storeAll operations.
    The remove operations need to be handled synchronously to avoid corruption of the data-set in the scenario of reading a entry from the cache immediately after removing it (if it were not synchronously deleted from the backing storage, then reading it back could give an incorrect non-null value). Therefore remove operations are handled synchronously on the service / worker thread, and not delayed on the write-behind thread.
    Gets are also synchronously handled, so they again are served on the service / worker thread.
    So if the puts are slow and wait too much, that may delay other puts but should not contend with other threads. If the puts are computation intensive, then obviously they hinder other threads because of consumption of the same CPU resource, and not simply because they execute.
    Best regards,
    Robert

  • Read-through/write-behind and queued deletes (and updates)

    Hi,
    If I am changing the state of objects in a cache and using the write-behind and read-through mechanism what happens when I have deleted or updated an object in the cache but the change has not yet been committed to the database?.
    If I delete and object in the cache and the delete DB operation is being queued and during this time try and perform a get against the key for the object is the value read through from the database or is it ignored since the database delete is pending?
    For updates I presume that the value in the cache will be used - as the value exists in the cache and a read-through from the database will not be triggered.
    Can you clarify the behavior of Coherence under these circumstances, particularly that of a pending delete.
    Thanks,
    Dave

    Hi Dave,
    If I am changing the state of objects in a cache and
    using the write-behind and read-through mechanism
    what happens when I have deleted or updated an object
    in the cache but the change has not yet been
    committed to the database?.
    If I delete and object in the cache and the delete DB
    operation is being queued and during this time try
    and perform a get against the key for the object is
    the value read through from the database or is it
    ignored since the database delete is pending?I seem to remember a forum post mentioning that the removes from a write-behind cache are performed synchronously (they are done as part of the backingMap.remove(key)) operation so even if there are were pending updates in the write-behind queue. If I remember correctly, then the above mentioned problem cannot happen.
    >
    For updates I presume that the value in the cache
    will be used - as the value exists in the cache and a
    read-through from the database will not be
    triggered.
    Exactly.
    Best regards,
    Robert

  • Coherence 3.3.1 Version, Write Behind Replicated Cache Error

    Hi,
    I am using Cohernece 3.3.1 Version, i have a Write Behind cache, the put method is throwing following exception:
    java.lang.IllegalArgumentException: Invalid internal format: Inactive
    at com.tangosol.coherence.component.util.BackingMapManagerContext.addInternalValueDecoration(BackingMapManagerCo
    ntext.CDB:11)
    at com.tangosol.net.cache.ReadWriteBackingMap.put(ReadWriteBackingMap.java:737)
    at com.tangosol.coherence.component.util.CacheHandler.onLeaseUpdate(CacheHandler.CDB:52)
    at com.tangosol.coherence.component.util.daemon.queueProcessor.service.ReplicatedCache.performUpdate(ReplicatedC
    ache.CDB:11)
    at com.tangosol.coherence.component.util.daemon.queueProcessor.service.ReplicatedCache.onLeaseUpdateRequest(Repl
    icatedCache.CDB:22)
    at com.tangosol.coherence.component.util.daemon.queueProcessor.service.ReplicatedCache$LeaseUpdateRequest.onRece
    ived(ReplicatedCache.CDB:5)
    at com.tangosol.coherence.component.util.daemon.queueProcessor.Service.onMessage(Service.CDB:9)
    at com.tangosol.coherence.component.util.daemon.queueProcessor.Service.onNotify(Service.CDB:123)
    at com.tangosol.coherence.component.util.daemon.queueProcessor.service.ReplicatedCache.onNotify(ReplicatedCache.
    CDB:3)
    at com.tangosol.coherence.component.util.Daemon.run(Daemon.CDB:35)
    at java.lang.Thread.run(Thread.java:534)
    ============
    The same cache works fine if i change the
    value of <write-delay-seconds> parameter to 0 i.e. if i make the cache write through.
    Could someone help me out with this issue.
    -thanks
    Krishan

    Write-behind caching is not supported with Replicated cache. Even with write-through, you'll end up generating replicated writes back to the back-end database, drastically increasing load.
    For more details, please see:
    http://wiki.tangosol.com/display/COH33UG/Read-Through,+Write-Through,+Refresh-Ahead+and+Write-Behind+Caching
    For applications where write-behind would be used, the partitioned (distributed) cache is almost always a far better option. Is there a reason to not use this?
    Jon Purdy
    Oracle

  • TTL specified in put operation doesn't always work when using write-behind

    I'm using a distributed cache with a write-behind cache store (see the config below). I found that when I do something like myCache.put(key, value, ttl), the entry survives the specified ttl. I tried doing the same with a distributed cache with a write-through cachestore and there everything does happen correctly.
    Is this sort of operation not permitted in caches containing a write-behind cachestore? If not wouldn't it be better to throw an UnsupportedOperationException.
    I created a small test to simulate this. I added values to the cache with a TTL of 1 to 10 seconds and found that the 10 second entries stayed in the cache.
    Configuration used:
    <?xml version="1.0"?>
    <!DOCTYPE cache-config SYSTEM "cache-config.dtd">
    <cache-config>
         <caching-scheme-mapping>
              <cache-mapping>
                   <cache-name>TTL_TEST</cache-name>
                   <scheme-name>testScheme</scheme-name>
              </cache-mapping>
         </caching-scheme-mapping>
         <caching-schemes>
              <distributed-scheme>
                   <scheme-name>testScheme</scheme-name>
                   <service-name>testService</service-name>
                   <backing-map-scheme>
                        <read-write-backing-map-scheme>
                             <internal-cache-scheme>
                                  <local-scheme>
                                       <service-name>testBackLocalService</service-name>
                                  </local-scheme>
                             </internal-cache-scheme>
                             <cachestore-scheme>
                                  <class-scheme>
                                       <scheme-name>testBackStore</scheme-name>
                                       <class-name>TTLTestServer$TestCacheStore</class-name>
                                  </class-scheme>
                             </cachestore-scheme>
                             <write-delay>3s</write-delay>
                        </read-write-backing-map-scheme>
                   </backing-map-scheme>
                   <local-storage>true</local-storage>
                   <autostart>true</autostart>
              </distributed-scheme>
         </caching-schemes>
    </cache-config>Code of test:
    import java.util.Collection;
    import java.util.List;
    import java.util.Map;
    import java.util.concurrent.Callable;
    import java.util.concurrent.ExecutionException;
    import java.util.concurrent.ExecutorService;
    import java.util.concurrent.Executors;
    import java.util.concurrent.Future;
    import org.joda.time.DateTime;
    import org.joda.time.Duration;
    import org.slf4j.Logger;
    import org.slf4j.LoggerFactory;
    import org.springframework.util.StopWatch;
    import org.testng.annotations.BeforeClass;
    import org.testng.annotations.Test;
    import com.google.common.collect.Lists;
    import com.tangosol.net.CacheFactory;
    import com.tangosol.net.NamedCache;
    import com.tangosol.net.cache.CacheStore;
    @Test
    public class TTLTestServer
         private static final int RETRIES = 5;
         private static final Logger logger = LoggerFactory.getLogger( TTLTestServer.class );
         private NamedCache m_cache;
          * List of Time-To-Lives in seconds to check
         private final List<Integer> m_listOfTTLs = Lists.newArrayList(1, 3, 5, 10);
          * Test is done in separate threads to speed up the test
         private final  ExecutorService m_executorService = Executors.newCachedThreadPool();
         @BeforeClass
         public void setup()
              logger.info("Getting the cache");
              m_cache =  CacheFactory.getCache("TTL_TEST");
         public static class TestCacheStore implements CacheStore
              public void erase(Object arg0)
              public void eraseAll(Collection arg0)
              public void store(Object arg0, Object arg1)
              public void storeAll(Map arg0)
              public Object load(Object arg0)
              {return null;}
              public Map loadAll(Collection arg0)
              {return null;}
         public void testTTL() throws InterruptedException, ExecutionException
              logger.info("Starting TTL test");
              List<Future<StopWatch>> futures = Lists.newArrayList();
              for (final Integer ttl : m_listOfTTLs)
                   futures.add(m_executorService.submit(new Callable()
                        public Object call() throws Exception
                             StopWatch stopWatch= new StopWatch("TTL=" + ttl);
                             for (int retry = 0; retry < RETRIES; retry++)
                                  logger.info("Adding a value in cache for TTL={} in try={}", ttl, retry+1);
                                  stopWatch.start("Retry="+retry);
                                  m_cache.put(ttl, null, ttl*1000);
                                  waitUntilNotInCacheAnymore(ttl, retry);
                                  stopWatch.stop();
                             return stopWatch;
                        private void waitUntilNotInCacheAnymore(final Integer ttl, final int currentTry) throws InterruptedException
                             DateTime startTime = new DateTime();
                             long maxMillisToWait = ttl*2*1000;     //wait max 2 times the time of the ttl
                             while(m_cache.containsKey(ttl) )
                                  Duration timeTaken = new Duration(startTime, new DateTime());
                                  if(timeTaken.getMillis() > maxMillisToWait)
                                       throw new RuntimeException("Already waiting " + timeTaken + " for ttl=" + ttl + " and retry=" +  currentTry);
                                  Thread.sleep(1000);
              logger.info("Waiting until all futures are finished");
              m_executorService.shutdown();
              logger.info("Getting results from futures");
              for (Future<StopWatch> future : futures)
                   StopWatch sw = future.get();
                   logger.info(sw.prettyPrint());
    }Failure message:
    FAILED: testTTL
    java.util.concurrent.ExecutionException: java.lang.RuntimeException: Already waiting PT20.031S for ttl=10 and retry=0
         at java.util.concurrent.FutureTask$Sync.innerGet(Unknown Source)
         at java.util.concurrent.FutureTask.get(Unknown Source)
         at TTLTestServer.testTTL(TTLTestServer.java:159)
    Caused by: java.lang.RuntimeException: Already waiting PT20.031S for ttl=10 and retry=0
         at TTLTestServer$1.waitUntilNotInCacheAnymore(TTLTestServer.java:139)
         at TTLTestServer$1.call(TTLTestServer.java:122)
         at java.util.concurrent.FutureTask$Sync.innerRun(Unknown Source)
         at java.util.concurrent.FutureTask.run(Unknown Source)
         at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(Unknown Source)
         at java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source)
         at java.lang.Thread.run(Unknown Source)I'm using Coherence 3.4.2.
    Best regards
    Jan

    Hi, still no luck. However, I noticed that setting the write-delay value of the write-behind store to 0s or 1s, solved the problem. It only starts to given me "the node has already been removed" excpetions once the write-delay value is 2s or higher.
    You can find the coherence-cache-config.xml below:
    <?xml version="1.0"?>
    <!DOCTYPE cache-config SYSTEM "cache-config.dtd">
    <cache-config>
         <caching-scheme-mapping>
              <cache-mapping>
                   <cache-name>TTL_TEST</cache-name>
                   <scheme-name>testScheme</scheme-name>
              </cache-mapping>
         </caching-scheme-mapping>
         <caching-schemes>
              <distributed-scheme>
                   <scheme-name>testScheme</scheme-name>
                   <service-name>testService</service-name>
                   <backing-map-scheme>
                        <read-write-backing-map-scheme>
                             <internal-cache-scheme>
                                  <local-scheme>
                                       <service-name>testBackLocalService</service-name>
                                  </local-scheme>
                             </internal-cache-scheme>
                             <cachestore-scheme>
                                  <class-scheme>
                                       <scheme-name>testBackStore</scheme-name>
                                       <class-name>TTLTestServer$TestCacheStore</class-name>
                                  </class-scheme>
                             </cachestore-scheme>
                             <write-delay>2s</write-delay>
                        </read-write-backing-map-scheme>
                   </backing-map-scheme>
                   <local-storage>true</local-storage>
                   <autostart>true</autostart>
              </distributed-scheme>
         </caching-schemes>
    </cache-config>You can find the test program below:
    import java.util.ArrayList;
    import java.util.Collection;
    import java.util.List;
    import java.util.Map;
    import org.joda.time.DateTime;
    import org.joda.time.Duration;
    import org.springframework.util.StopWatch;
    import com.tangosol.net.CacheFactory;
    import com.tangosol.net.NamedCache;
    import com.tangosol.net.cache.CacheStore;
    public class TTLTestServer
         private static final int RETRIES = 5;
         private NamedCache m_cache;
          * List of Time-To-Lives in seconds to check
         private final List<Integer> m_listOfTTLs = new ArrayList<Integer>();
          * @param args
          * @throws Exception
         public static void main( String[] args ) throws Exception
              new TTLTestServer().test();
          * Empty CacheStore
          * @author jbe
         public static class TestCacheStore implements CacheStore
              public void erase(Object arg0)
              @SuppressWarnings ( "unchecked" )
              public void eraseAll(Collection arg0)
              public void store(Object arg0, Object arg1)
              @SuppressWarnings ( "unchecked" )
              public void storeAll(Map arg0)
              public Object load(Object arg0)
              {return null;}
              @SuppressWarnings ( "unchecked" )
              public Map loadAll(Collection arg0)
              {return null;}
          * Sets up and executes the test setting values in a cache with a given time-to-live value and waiting for the value to disappear.
          * @throws Exception
         private void test() throws Exception
              System.out.println(new DateTime() + " - Setting up TTL test");
              m_cache =  CacheFactory.getCache("TTL_TEST");
              m_listOfTTLs.add( 1 );
              m_listOfTTLs.add( 3 );
              m_listOfTTLs.add( 5 );
              m_listOfTTLs.add( 10);
              System.out.println(new DateTime() + " - Starting TTL test");
              for (final Integer ttl : m_listOfTTLs)
                   StopWatch sw = doTest(ttl);
                   System.out.println(sw.prettyPrint());
          * Adds a value to the cache with the time-to-live as given by the ttl parameter and waits until it's removed from the cache.
          * Repeats this {@link #RETRIES} times
          * @param ttl
          * @return
          * @throws Exception
         private StopWatch doTest(Integer ttl) throws Exception
              StopWatch stopWatch= new StopWatch("TTL=" + ttl);
              for (int retry = 0; retry < RETRIES; retry++)
                   System.out.println(new DateTime() + " - Adding a value in cache for TTL=" + ttl + " in try= " + (retry+1));
                   stopWatch.start("Retry="+retry);
                   m_cache.put(ttl, null, ttl*1000);
                   waitUntilNotInCacheAnymore(ttl, retry);
                   stopWatch.stop();
              return stopWatch;
          * Wait until the value for the given ttl is not in the cache anymore
          * @param ttl
          * @param currentTry
          * @throws InterruptedException
         private void waitUntilNotInCacheAnymore(final Integer ttl, final int currentTry) throws InterruptedException
              DateTime startTime = new DateTime();
              long maxMillisToWait = ttl*2*1000;     //wait max 2 times the time of the ttl
              while(m_cache.containsKey(ttl) )
                   Duration timeTaken = new Duration(startTime, new DateTime());
                   if(timeTaken.getMillis() > maxMillisToWait)
                        throw new RuntimeException("Already waiting " + timeTaken + " for ttl=" + ttl + " and retry=" +  currentTry);
                   Thread.sleep(1000);
    }You can find the output below:
    2009-12-03T11:50:04.584+01:00 - Setting up TTL test
    2009-12-03 11:50:04.803/0.250 Oracle Coherence 3.5.2/463p2 <Info> (thread=main, member=n/a): Loaded operational configuration from resource "jar:file:/C:/Temp/coherence3.5.2/coherence-java-v3.5.2b463-p1_2/coherence/lib/coherence.jar!/tangosol-coherence.xml"
    2009-12-03 11:50:04.803/0.250 Oracle Coherence 3.5.2/463p2 <Info> (thread=main, member=n/a): Loaded operational overrides from resource "jar:file:/C:/Temp/coherence3.5.2/coherence-java-v3.5.2b463-p1_2/coherence/lib/coherence.jar!/tangosol-coherence-override-dev.xml"
    2009-12-03 11:50:04.803/0.250 Oracle Coherence 3.5.2/463p2 <D5> (thread=main, member=n/a): Optional configuration override "/tangosol-coherence-override.xml" is not specified
    2009-12-03 11:50:04.803/0.250 Oracle Coherence 3.5.2/463p2 <D5> (thread=main, member=n/a): Optional configuration override "/custom-mbeans.xml" is not specified
    Oracle Coherence Version 3.5.2/463p2
    Grid Edition: Development mode
    Copyright (c) 2000, 2009, Oracle and/or its affiliates. All rights reserved.
    2009-12-03 11:50:04.943/0.390 Oracle Coherence GE 3.5.2/463p2 <Info> (thread=main, member=n/a): Loaded cache configuration from "file:/C:/jb/workspace3.5/TTLTest/target/classes/coherence-cache-config.xml"
    2009-12-03 11:50:05.318/0.765 Oracle Coherence GE 3.5.2/463p2 <D5> (thread=Cluster, member=n/a): Service Cluster joined the cluster with senior service member n/a
    2009-12-03 11:50:08.568/4.015 Oracle Coherence GE 3.5.2/463p2 <Info> (thread=Cluster, member=n/a): Created a new cluster "cluster:0xD3FB" with Member(Id=1, Timestamp=2009-12-03 11:50:05.193, Address=172.16.44.32:8088, MachineId=36896, Location=process:11848, Role=TTLTestServerTTLTestServer, Edition=Grid Edition, Mode=Development, CpuCount=2, SocketCount=2) UID=0xAC102C20000001255429380990201F98
    2009-12-03 11:50:08.584/4.031 Oracle Coherence GE 3.5.2/463p2 <D5> (thread=Invocation:Management, member=1): Service Management joined the cluster with senior service member 1
    2009-12-03 11:50:08.756/4.203 Oracle Coherence GE 3.5.2/463p2 <D5> (thread=DistributedCache:testService, member=1): Service testService joined the cluster with senior service member 1
    2009-12-03T11:50:08.803+01:00 - Starting TTL test
    2009-12-03T11:50:08.818+01:00 - Adding a value in cache for TTL=1 in try= 1
    2009-12-03T11:50:09.818+01:00 - Adding a value in cache for TTL=1 in try= 2
    Exception in thread "main" (Wrapped: Failed request execution for testService service on Member(Id=1, Timestamp=2009-12-03 11:50:05.193, Address=172.16.44.32:8088, MachineId=36896, Location=process:11848, Role=TTLTestServerTTLTestServer)) java.lang.IllegalStateException: the node has already been removed
         at com.tangosol.util.Base.ensureRuntimeException(Base.java:293)
         at com.tangosol.coherence.component.util.daemon.queueProcessor.service.Grid.tagException(Grid.CDB:36)
         at com.tangosol.coherence.component.util.daemon.queueProcessor.service.grid.DistributedCache.onContainsKeyRequest(DistributedCache.CDB:41)
         at com.tangosol.coherence.component.util.daemon.queueProcessor.service.grid.DistributedCache$ContainsKeyRequest.run(DistributedCache.CDB:1)
         at com.tangosol.coherence.component.net.message.requestMessage.DistributedCacheKeyRequest.onReceived(DistributedCacheKeyRequest.CDB:12)
         at com.tangosol.coherence.component.util.daemon.queueProcessor.service.Grid.onMessage(Grid.CDB:9)
         at com.tangosol.coherence.component.util.daemon.queueProcessor.service.Grid.onNotify(Grid.CDB:136)
         at com.tangosol.coherence.component.util.daemon.queueProcessor.service.grid.DistributedCache.onNotify(DistributedCache.CDB:3)
         at com.tangosol.coherence.component.util.Daemon.run(Daemon.CDB:42)
         at java.lang.Thread.run(Thread.java:619)
    Caused by: java.lang.IllegalStateException: the node has already been removed
         at com.tangosol.util.AbstractSparseArray$Crawler.remove(AbstractSparseArray.java:1274)
         at com.tangosol.net.cache.OldCache.evict(OldCache.java:580)
         at com.tangosol.net.cache.OldCache.containsKey(OldCache.java:171)
         at com.tangosol.net.cache.ReadWriteBackingMap.containsKey(ReadWriteBackingMap.java:597)
         at com.tangosol.coherence.component.util.daemon.queueProcessor.service.grid.DistributedCache.onContainsKeyRequest(DistributedCache.CDB:25)
         ... 7 more
    2009-12-03 11:50:10.834/6.281 Oracle Coherence GE 3.5.2/463p2 <D4> (thread=ShutdownHook, member=1): ShutdownHook: stopping cluster node
    2009-12-03 11:50:10.834/6.281 Oracle Coherence GE 3.5.2/463p2 <D5> (thread=Cluster, member=1): Service Cluster left the clusterBest regards
    Jan

Maybe you are looking for