TODO

   1 Redis TODO
   2 ----------
   3
   4 WARNING: are you a possible Redis contributor?
   5          Before implementing what is listed what is listed in this file
   6          please drop a message in the Redis google group or chat with
   7          antirez or pietern on irc.freenode.org #redis to check if the work
   8          is already in progress and if the feature is still interesting for
   9          us, and *how* exactly this can be implemented to have good changes
  10          of a merge. Otherwise it is probably wasted work! Thank you
  11
  12
  13 API CHANGES
  14 ===========
  15
  16 * Turn commands into variadic versions when it makes sense, that is, when
  17   the variable number of arguments represent values, and there is no conflict
  18   with the return value of the command.
  19
  20 CLUSTER
  21 =======
  22
  23 * Implement rehashing and cluster check in redis-trib.
  24 * Reimplement MIGRATE / RESTORE to use just in memory buffers (no disk at
  25   all). This will require touching a lot of the RDB stuff around, but we may
  26   hand with faster persistence for RDB.
  27 * Implement the slave nodes semantics and election.
  28 * Allow redis-trib to create a cluster-wide snapshot (using SYNC).
  29 * Allow redis-trib to restore a cluster-wide snapshot (implement UPLOAD?).
  30
  31 APPEND ONLY FILE
  32 ================
  33
  34 * in AOF rewirte use HMSET to rewrite small hashes instead of multiple calls
  35   to HSET.
  36
  37 OPTIMIZATIONS
  38 =============
  39
  40 * Avoid COW due to incrementing the dict iterators counter.
  41 * SORT: Don't copy the list into a vector when BY argument is constant.
  42 * Write the hash table size of every db in the dump, so that Redis can resize the hash table just one time when loading a big DB.
  43 * Read-only mode for slaves.
  44 * Redis big lists as linked lists of small ziplists?
  45   Possibly a simple heuristic that join near nodes when some node gets smaller than the low_level, and split it into two if gets bigger than high_level.
  46
  47 REPORTING
  48 =========
  49
  50 * Better INFO output with sections.
  51
  52 RANDOM
  53 ======
  54
  55 * Clients should be closed as far as the output buffer list is bigger than a given number of elements (configurable in redis.conf)
  56 * Should the redis default configuration, and the default redis.conf, just bind 127.0.0.1?
  57
  58 KNOWN BUGS
  59 ==========
  60
  61 * What happens in the following scenario:
  62     1) We are reading an AOF file.
  63     2) SETEX FOO 5 BAR
  64     3) APPEND FOO ZAP
  65     What happens if between 1 and 2 for some reason (system under huge load
  66     or alike) too many time passes? We should prevent expires while the
  67     AOF is loading.
  68 * #519: Slave may have expired keys that were never read in the master (so a DEL
  69   is not sent in the replication channel) but are already expired since
  70   a lot of time. Maybe after a given delay that is undoubltly greater than
  71   the replication link latency we should expire this key on the slave on
  72   access?
  73
  74 DISKSTORE TODO
  75 ==============
  76
  77 * Fix FLUSHALL/FLUSHDB: the queue of pending reads/writes should be handled.
  78 * Check that 00/00 and ff/ff exist at startup, otherwise exit with error.
  79 * Implement sync flush option, where data is written synchronously on disk when a command is executed.
  80 * Implement MULTI/EXEC as transaction abstract API to diskstore.c, with transaction_start, transaction_end, and a journal to recover.
  81 * Stop BGSAVE thread on shutdown and any other condition where the child is killed during normal bgsave.
  82 * Fix RANDOMKEY to really do something interesting
  83 * Fix DBSIZE to really do something interesting
  84 * Add a DEBUG command to check if an entry is or not in memory currently
  85 * dscache.c near 236, kobj = createStringObject... we could use static obj.