Redis TODO and Roadmap
+----------------------
-VERSION 1.4 TODO (Hash type)
-============================
-
-* BRPOPLPUSH
-* RPOPLPUSH should notify blocking POP operations
-* List ops like L/RPUSH L/RPOP should return the new list length.
-* Save dataset / fsync() on SIGTERM
-* MULTI/EXEC should support the "EXEC FSYNC" form
-* Synchronous Virtual Memory
-* BLPOP & C. tests (write a non blocking Tcl client as first step)
-
-Virtual Memory sub-TODO:
-* Check if the page selection algorithm is working well.
-* Divide swappability of objects by refcount
-* vm-swap-file <filename>. The swap file should go where the user wants, and if it's already there and of the right size we can avoid to create it again.
-* it should be possible to give the vm-max-memory option in megabyte, gigabyte, ..., just using 2GB, 100MB, and so forth.
-* Make sure to wait all the IO threads are done before to fork() for BGSAVE and BGREWRITEAOF
-* Enlarge the stack of threads, to problems when calling LZF lib.
-
-VERSION 1.6 TODO (Virtual memory)
-=================================
-
-* Asynchronous Virtual Memory
-* Hashes (HSET, HGET, HEXISTS, HLEN, ...).
-
-VERSION 1.8 TODO (Fault tollerant sharding)
-===========================================
-
-* Redis-cluster, a fast intermediate layer (proxy) that implements consistent hashing and fault tollerant nodes handling.
-
-Interesting readings about this:
-
- - http://ayende.com/Blog/archive/2009/04/06/designing-rhino-dht-a-fault-tolerant-dynamically-distributed-hash.aspx
-
-VERSION 2.0 TODO (Optimizations and latency)
+VERSION 2.2 TODO (Optimizations and latency)
============================================
* Lower the CPU usage.
* Lower the RAM usage everywhere possible.
-* Use epool and alike to rewrite ae.c for Linux and other platforms suppporting fater-than-select() mutiplexing APIs.
-* Implement an UDP interface for low-latency GET/SET operations.
+* Specially encoded Sets (like Hashes).
+* Implement an UDP interface for low-latency operations.
+* What about a special coding that is about storing the "rdb" serialized format instead of the actual value? This can be used when we have LRU in order to super-compress data into memory, for data not accessed frequetly. It's a VM-alike strategy but fully in memory, may reduce the space to hold some dataset in an impressive way. Trivial to implement.
+* Another idea: LRU does not need to be super precise right? Maybe it's a good idea to just put into the skiplist implementing the LRU just the pointer to the key without evne incr/decr business, nor the need to remove the pointer when the key is deleted. There is to think more about that.
-VERSION 2.2 TODO (Optimizations and latency)
-============================================
+VERSION 2.x TODO
+================
-* JSON command able to access data serialized in JSON format. For instance if I've a key foobar with a json object I can alter the "name" file using somthing like: "JSON SET foobar name Kevin". We should have GET and INCRBY as well.
+* BRPOPLPUSH
+* Save dataset / fsync() on SIGTERM
+* Change the implementation of ZCOUNT to use the augmented skiplist in order to be much faster.
+
+Virtual Memory optimizations:
+* Use multiple open FDs against the VM file, one for thread.
+* Check what happens performance-wise if instead of creating threads again and again the same threads are reused forever. Note: this requires a way to disable this clients in the child, but waiting for empty new jobs queue can be enough.
+* Implement LEN, PEEK, POKE, SETBIT, GETBIT
OTHER IMPORTANT THINGS THAT WILL BE ADDED BUT I'M NOT SURE WHEN
===============================================================
* Specially encoded memory-saving integer sets.
* A command to export a JSON dump (there should be mostly working patch needing major reworking).
* Specially encoded sets of integers (this includes a big refactoring providing an higher level layer for Sets manipulation)
-* ZRANK: http://docs.google.com/viewer?a=v&q=cache:tCQaP3ZeN4YJ:courses.csail.mit.edu/6.046/spring04/handouts/ps5-sol.pdf+skip+list+rank+operation+augmented&hl=en&pid=bl&srcid=ADGEEShXuNjTcZyXw_1cq9OaWpSXy3PprjXqVzmM-LE0ETFznLyrDXJKQ_mBPNT10R8ErkoiXD9JbMw_FaoHmOA4yoGVrA7tZWiy393JwfCwuewuP93sjbkzZ_gnEp83jYhPYjThaIzw&sig=AHIEtbRF0GkYCdYRFtTJBE69senXZwFY0w
SMALL ONES:
+* If sizeof(double) == sizeof(void*) we could store the double value of sorted sets directly in place of the pointer instead of allocating it in the heap.
+* Delete on writes against expire policy should only happen after argument parsing for commands doing their own arg parsing stuff.
* Give errors when incrementing a key that does not look like an integer, when providing as a sorted set score something can't be parsed as a double, and so forth.
* MSADD (n keys) (n values). See this thread in the Redis google group: http://groups.google.com/group/redis-db/browse_thread/thread/e766d84eb375cd41
* Don't save empty lists / sets / zsets on disk with snapshotting.
* Remove keys when a list / set / zset reaches length of 0.
+* An option to exec a command slave-side if the master connection is lost: even cooler: if the script returns "0" the slave elects itself as master, otherwise continue trying to reconnect.
+* PING the master from time to time to check if it's gone.
THE "MAYBE" TODO LIST: things that may or may not get implemented
=================================================================
* Pattern-matching replication.
* Add an option to relax the delete-expiring-keys-on-write semantic *denying* replication and AOF when this is on? Can be handy sometimes, when using Redis for non persistent state, but can create problems. For instance should rename and move also "move" the timeouts? How does this affect other commands?
* Multiple BY in SORT.
+
+KNOWN BUGS
+==========
+
+* LRANGE and other commands are using 32 bit integers for ranges, and overflows are not detected. So LRANGE mylist 0 23498204823094823904823904 will have random effects.
projects / redis.git / blobdiff