Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Hemraj Kerkar
    @reoxey
    Hi, Is there any nodejs package for KeyDB? Thanks
    ibsusu
    @ibsusu
    hello I'm wondering about using AOF and RDB along with the scratch file
    is there something that could tell me whether I should use them all at once or whether using aof and the scratch file is enough
    Stuart Williams (CTT)
    @stuwilliams-clgx
    It would be nice if there was a place where we could download just the CLI. I looked around and did not see one? Is there one?
    Wasim Baig
    @wasimbaig
    Hi guys, using keydb to replace a redis installation. My requirement is to have 4 keydb on 2 vm, all replicated (but I can live with writing on only one keydb of each vm, ideally both). So I setup 2 VMs, A and B. A has two KeyDB on it A1 and A2, and same for B(B1, B2). Active-active between A1:B1. And, A2 is replicaof A1, and B2 is replicaof B1. SET testkey 123 on A1, its replicated on A2 and B1. But not on B2. What am I missing here?
    Greg TAPPERO
    @coulix
    Hello there, I migrated redis to keydb, all is working as expected. I had one question, is keydb going to continue working when I reach the max memory available (maxmemory) ? Is there anyway to make it work with disk instead of ram.
    Werner Barnard
    @wernerbarnard_twitter
    Haai - are there perhaps any more detailed docs on running active-active using docker swarm and auto-scaling, i.e. the configuration is linked to all containers under a swarm service, not specific IP addresses (as those might change) ?
    Sam Partee
    @Spartee
    Hi all! I have been looking into KeyDB and I’m wondering if there has been anyone who has tried this in an HPC setting before? aka launching through a workload manager like Slurm? Ive been very impressed by some initial testing
    Sam Partee
    @Spartee
    also, what is the preferred method of creating a keyDB cluster accross multiple hosts?
    through the cli? our own version of the create_cluster script?(seemingly supposed to be for testing)
    RNZ
    @rnz
    hi!
    Why i see in keydb-benchmark ouput: "multi-thread: no"?
    Roy
    @roynasser
    Hi all, we are considering using keydb as a replacement for our current redis implementation (as cache / kv store / job queue - different clusters for each type already). As we are running in kubernetes, I'd like to know if there is any documentation about running keydb on kubernetes, etc... Would using a redis chart or operator work just by switching the redis image with a keydb one?
    Assaf
    @AssafPeer_twitter
    Hi, This is a great project.
    Few questions related to using Flash:
    1. Did you have a chance to look into the SSD durability and excessive write IOs to the drives? 2. Do you have some documentation I can refer to in order to define RAM vs. SSD memory utilization and how to define their sizes? 3) Any available benchmark results with limited RAM and NVMe SSD?
    Assaf
    @AssafPeer_twitter

    Hello all,
    Reading the KeyDB flash-storage document, it is stated that both zfs and btrfs are supported. However, when trying to use zfs I get the following error
    "# Scratch file system does not support Copy on Write. To fix this scratch-file-path must point to a path on a filesystem which supports copy on write, such as btrfs."

    Is zfs still supported? Any idea what could have gone wrong?
    thanks!

    Joe Mordica
    @jmordica
    Hi there! Any CentOS 6/7 RPM's lingering anywhere out there?
    of KeyDB of course :)
    ronen-cedato
    @ronen-cedato
    Hi, a question regarding active replica split brain issue. When I am using heavy load on 2 keydb instances in active-active mode (~1000 requests per second), I get a split brain scenario. All my requests increment the same entry in a hash. When I try to read the value from the hash key, I get different result than I expect, like some requests were discarded so the value I get is lower from what I would expect. Any known issue about it? Any existing benchmark for heavy load as a reference?
    Assaf
    @AssafPeer_twitter
    I am running a "populate only" test with 1 KB objects and a 1:5 RAM to Flash capacity ratio and results are very low and unstable. I am using BTRFS and a Samsung 960 Pro NVMe device. Does anyone have current performance results when working with Flash?
    Paul Chen
    @paulmchen
    Do you have the detailed test info, for example the total number of requests sent, the test tool used, and the test tool parameters sent? One issue we saw is the throughput is significantly coming down after loading more than 280 GB of data to the Flash. This is running against the release 5 version from the community GitHub repo.
    Assaf
    @AssafPeer_twitter

    @paulmchen I am using memtier, both client and server are on the same machine with 32 GB of RAM and a Samsung 960 PRO 512 GB SSD. I am writing 50M 1000B objects using the following command:
    memtier_benchmark -s 127.0.0.1 -p 6379 -d 1000 --key-pattern=P:P -n allkeys --ratio=1:0 --key-maximum=50000000 --distinct-client-seed

    Initially the test runs at ~300 MB/s but after a few seconds when the RAM gets full, I start seeing significant fluctuations in performance and it even drops to the range of a few MB/s
    KeyDB server is running with 2 threads.

    John Sully
    @JohnSully
    FLASH is implemented with a large memory mapped file. The expectation was the OS would do a reasonable job paging but that hasn’t worked out in practice.
    We’re rearchitecting this feature to handle this explicitly along with a bunch of other architectural improvements including forkless background save
    The forked background save is the reason you need BTRFS right now
    That said there are kernel tuning parameters available for the current implementation. In particular swappiness and
    Page cache tuning parameters
    Yahook
    @Yahook
    Hello guys, anyone has experience running KeyDB on flash?
    Yahook
    @Yahook
    My amount of data is bigger that can hold RAM, I want to buy SSD 4TB for example and store 70-80% of cold data there and store in ram only hot data 20-30%. I also want to limit memory usase manually for KeyDB. For example let it only use 32GB of RAM. Is it possible to do that with KeyDB? Are there any results of performance tests? Right now I'm using Redis and I'm very unhappy that it stores evertyhing in RAM.
    hezhiqiang
    @hezhiqiang
    when we have multiple keydb host, can we use redisson lock?
    John Sully
    @JohnSully
    @hezhiqiang Yes it should be compatible with Redisson although I haven’t tested it specifically. KeyDB has a goal of 100% Redis compatibility.
    @Yahook Performance of the current FLASH feature drops off pretty quickly as the dataset exceeds RAM. We have a new version of FLASH coming out for January that is much faster.
    hezhiqiang
    @hezhiqiang
    can keydb be a slave of redis?
    hezhiqiang
    @hezhiqiang
    when keydb is active-active mode, which one should redisson user: replicated-servers, master-slave-servers, cluster-servers?
    hezhiqiang
    @hezhiqiang
    when i make keydb in centos, show this error: unrecognized command line option ‘-std=c++14’
    hezhiqiang
    @hezhiqiang
    i installed gcc 4.9.2, std=c++14 error is solved, but now is uuid/uuid.h: No such file or directory
    hezhiqiang
    @hezhiqiang
    now is this error: compactvector.h:18:35: error: ‘is_trivially_copyable’ is not a member of ‘std’
    static_assert(MEMMOVE_SAFE || std::is_trivially_copyable<T>::value, "compactvector requires trivially copyable types");
    which version of gcc shoud i use?
    hezhiqiang
    @hezhiqiang
    now make is sucessful, but when make test, has this error:
    * [err]: BITOP lshift string in tests/unit/bitops.tcl
    Expected 'ïg' to equal or match 'abcdefg'
    hezhiqiang
    @hezhiqiang
    haha, i installed ok, but when i set keydb be a slave of redis, has this error: non-fatal: Master doesn't understand REPLCONF uuid
    张舒鳌
    @zhangshuao
    hi gays
    @hezhiqiang yum -y install uuid-devel can resolve it .(i installed gcc 4.9.2, std=c++14 error is solved, but now is uuid/uuid.h: No such file or directory)
    张舒鳌
    @zhangshuao
    @JohnSully Can write an installation document about CentOS 7 in Wiki ?
    张舒鳌
    @zhangshuao
    make MALLOC=memkind
    John Sully
    @JohnSully
    Yes KeyDB can replicate from Redis
    sorry about the compile issues your having. Almost everyone is pulling from docker these days so thats what we focus on.
    Please open a bug to track a CentOS build instructions document and I’ll get to it when I have a chance
    张舒鳌
    @zhangshuao
    @JohnSully This is my issue: JohnSully/KeyDB#118
    make MALLOC=memkind (flash support)