jankotek/mapdb

MapDB provides concurrent Maps, Sets and Queues backed by disk storage or off-heap-memory. It is a fast and easy to use embedded Java database engine.

Markus Döring

@mdoering

Since I updated mapdb from 3.0.5 to 3.0.7 Intellij is bogus, unable to run any tests (if you try just nothing happens and the run icon is greed out forever). Is that known?

I see Kotlin versions has been updated, could that clash?
https://github.com/jankotek/mapdb/compare/mapdb-3.0.5...mapdb-3.0.7

Joshua Chambers

@jmclulus

Hey, I'm trying to get mapDB 3.0.7 working, I've included it via maven, but when I try to follow these examples, the classes it references just can't be found! I can browse in the eclipse "Maven Dependencies" and see the mapdb-3.0.7.jar file there, and I can browse it and see org.mapdb and then I can browse that and see the DB class... but I can't import the DB class! It keeps not being able to find it!

Joshua Chambers

@jmclulus

Hmm... 3.0.5 appears to resolve just fine though.

Aleksandr Bezhan

@alexbezhan

Hey guys. I'm looking into maybe migrating my app from MySQL to MapDB, because I've got sick of migrations and SQL stuff.

My question is how is data migration handled with MapDB ? Let's say I changed my class structure - added or removed a field ? How to deal with it ? Should I make IFs in serializer and check for fields with reflection or what ?

Gary Fry

@cdluv

Hi @here - I'm using the latest version of mapDB, and I've read in the documentation that there's a DB.compact() method available - however I cannot find this method in the source code at all! Does anyone know why it's not implemented in 3.0.7?

Gary Fry

@cdluv

@alexbezhan - it depends on how you store your values - if the the values are stored as JSON, then you can let the deserialiser worry about that for you.. any properties no longer required will simply not end up in your domain object, providing you use your correct annotation. If you are storing values in an internal format, then you'll need to have lots of IFs. There are ways to avoid this such as reorganising your database periodically - by migrating an old format to a new format (excluding unnecessary fields), but that would probably involve some downtime. There are ways, it just depends on how lazy you want to be!

ashaychangwani

@ashaychangwani

Hi, I'm very new here and I'm trying to create a Map that's stored on disk since my RAM is limited and the map is quite big. I've been trying to create a fileDB, and the size of the fileDB is set at 2048 kb and the RAM usage is increasing

Is there any fix?

Andreas Petersson

@apetersson

tried to shutdown a mmaped file and got: java.lang.NoSuchMethodError: sun.nio.ch.DirectBuffer.cleaner()Lsun/misc/Cleaner;

Per-Åke Minborg

@minborg

Have you checked out jankotek/mapdb#879 ?

Andreas Petersson

@apetersson

yes, but i would really love to cleanup the DB in my Junit tests cleanly, including truly deleting my db file. im on Java 11. disabling FileMmap is also not really an option

Per-Åke Minborg

@minborg

Since Java 9, the Cleaner was moved. It is no longer a part of sun.nio.ch.DirectBuffer. I am unsure how this is handled by mapdb? Anyone?

Andreas Petersson

@apetersson

is ti possible to set a different, fixed expiration dates for a group of keys/values?

Andreas Petersson

@apetersson

i implemented a FixedSizeSerializer for byte arrays where the size is known, and which skips writing the out.packInt(value.length); as in SerializerByteArray. for unknown reasons this is ca 50% slower in my impl...

Andreas Petersson

@apetersson

in this doc: https://studylib.net/doc/8397568/mapdb-cheat-sheet there is talks of using queues with mapdb. did not find anything in 3.0.7 was this removed? what to use instead when i want to persist a queue locally?

Jan Kotek

@jankotek

@apetersson Cheat Sheet was targeting older version (1.0x), there is no cheat sheet for new version.

Sergey Zharikhin

@SergeyZharikhin

hello, Guys! Is it possible to expire entries from inMemory db to onDisk db ONLY based on inMemory map store size?

Jan Kotek

@jankotek

@SergeyZharikhin I do not think so, not in newest version

Kevin Stiehl

@kstiehl

Hey gusy i am using only a memoryDirectDB at the moment to create a HTreeMap which then is accessed by mutliple threads. Is it necessary to call db.commit when I only use a the direct Memory. From the documentaion it looks like it is used to sync with the filesystem.

axonix

@axonix69_gitlab

after correct a fileDB commiting and closing , a database file remains locked until JVM shutdown on Windows10.
is there any way to force unlock it to make possible the file deleting ?

janothan

@janothan

What is the reason for a String Array Serializer not being available? What should I ideally choose for key -> String[]?

Jan Kotek

@jankotek

@kstiehl no it is not necessary, transactions are disabled by default, so that is fine

@axonix69_gitlab DBMaker.cleanerHackEnable(), it is JVM problem

Jan Kotek

@jankotek

@janothan I will add it in 4.0 jankotek/mapdb#937

Jasneet Singh

@singhjasneet

Hey @jankotek I am trying to fix the the cleanHackEnable() to work for Java 9+. But I am not able to build the project on my machine.

lexaz777

@lexaz777

Hello, I'm started to use mapdb 3.0.7, but there is no "db.compact()" method. Can you help me, please?

lexaz777

@lexaz777

@cdluv Did you find .compact() method?

sunyaf

@sunyaf

hello

java.lang.ArrayIndexOutOfBoundsException: 31 at org.mapdb.DataInput2$ByteArray.unpackLongArray(DataInput2.java:200)

this issue，any one konw

@jankotek

hello

any one can help me

jankotek/mapdb#948

@tokuhirom

can you help me?

sunyaf

@sunyaf

are you sleep?

satheessh

@satheessh

We need help on below error

java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)-->    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)-->    at io.netty.util.concurrent.FastThreadLocalRunnable.run(FastThreadLocalRunnable.java:30)-->    at java.lang.Thread.run(Thread.java:748)-->Caused by: java.lang.IllegalAccessError: Store was closed-->    at org.mapdb.StoreDirectAbstract.assertNotClosed(StoreDirectAbstract.kt:62)-->    at org.mapdb.StoreDirect.get(StoreDirect.kt:520)-->    at org.mapdb.BTreeMap.getNode(BTreeMap.kt:800)-->    at org.mapdb.BTreeMap.access$getNode(BTreeMap.kt:72)-->    at org.mapdb.BTreeMap$BTreeIterator.advanceFrom(BTreeMap.kt:1031)-->    at org.mapdb.BTreeMap$BTreeIterator.<init>(BTreeMap.kt:1016)-->    at org.mapdb.BTreeMap$valueIterator$1.<init>(BTreeMap.kt:1223)-->    at org.mapdb.BTreeMap.valueIterator(BTreeMap.kt:1223)-->    at org.mapdb.BTreeMap$values$1.iterator(BTreeMap.kt:999)-->    at java.util.Spliterators$IteratorSpliterator.estimateSize(Spliterators.java:1821)-->

Abhijith V Mohan

@avmohan

does the current stable 3.x have a concept of a FIFO queue? I don't see it anywhere in the docs, hence asking

star-usman

@star-usman

Hi, When we can expect version 4 release?

with HDFS storage integration

Yashwanth

@YashwanthKumarJ

Hi, While i was exploring on java offline data store, i came across mapdb. The requirement is to have a persistable queue. I tried building the latest codebase and use it in my project. Looks like queue is not working in FIFO basis.
File f = new File("queueTest");
DB db = DB.Maker.Companion.appendFile(f).make();
Queue<String> queue = QueueMaker.newLinkedFifoQueue(db, "myqueue", String.class).make();
queue.add("first");
queue.add("second");
queue.add("third");
queue.add("fourth");

System.out.println(queue);
System.out.println(queue.poll());

output

[fourth, third, second, first]
fourth

1) This is not FIFO queue result. I know that it is not officially released, but just in case if the implementation is in progress, please do take care of it. Also, if i have made any mistake in the way the queue is supposed to be created and used, do let me know.
2) I know in 3.X queues can be created. Is this queue persistable?

Abhijith V Mohan

@avmohan

@YashwanthKumarJ I had a similar requirement

Evaluated mapdb, square/tape, and openhft/chronicle-queue. Settled on chronicle-queue

I'd recommend you to take a look at both. Tape's api is much simpler, but if your use case needs high throughput, especially with good number of concurrent writers, cq will be better.

Marko Mitić

@mimarko

Hi all, any experience or best practice to use MapDB (file stored HTreeMap or hybrid - mem+file storage) with AWS Elastic Beanstalk and EC2 instances (could be scalable)? Thanks in advance!

Avi Hayun

@Chaiavi

Hi,

I am using mapdb v3.0.7

How do I start a simple queue, no need for any fancy features, just a simple mapdb implementation

If it is a problem in this version, I can change the version to any other one

Where communities thrive

People

Repo info

Activity

output