I also noticed some PRs that seem to be related (e.g. Netflix/dyno-queues#82) and was wondering if those were attempts to fix this issue
I am trying to build a dynomite-manager to make it work in my own data center. I could not make it work. And the instructions are not clear its very confusing or not accurate information. Does anyone made it work Dynomite manager working in our own hosted environment instead of aws ?
Basically i have 2 DC in 2 servers .if serverB goes down im able to set data to cluster using serverA redis nodes .when serverB redis nodes comes up ,im not able to get latest data using serverB nodes .any body can help me with this ? link : Netflix/dynomite#770
How can debug further?
what is the vnode in the below documentation taken from https://github.com/Netflix/dynomite
tokens: The token(s) owned by a node. Currently, we don't support vnode yet so this only works with one token for the time being.
tokens: The token(s) owned by a node. Currently, we don't support vnode yet so this only works with one token for the time being.
hi
while compiling the project with Jedis and DynoClient Jedis. It is giving compilation error for Jedis client related class though the required jar file is present.
Any suggestion for resolving the compilation issue.
Netflix/dyno#298
I am using JAVA8, JEDIS and MAVEN
Pl. let me know any other information required.
while compiling the project with Jedis and DynoClient Jedis. It is giving compilation error for Jedis client related class though the required jar file is present.
Any suggestion for resolving the compilation issue.
Netflix/dyno#298
I am using JAVA8, JEDIS and MAVEN
Pl. let me know any other information required.
can someone elaborate on the status of the project? is it still used by netflix etc ?
Hi, the project is still active. There will be more active development on it continuing soon. For now, there are some bandwidth issues on being able to reply to all OSS questions, we're hoping the community will be able to fill in here. But we're having one more developer join soon and keep the project more active.
Is this a small operation within the netflix org or is it pretty core
i am looking at it as a cassandra replacement for relatively small read heavy workloads
as cassandra tens to be "operationally complex"
@jelmerk Certain use-cases can use it as a primary datastore depending on the guarantees required. You can vet it based on the guarantees mentioned in the wiki here:
https://github.com/Netflix/dynomite/wiki
https://github.com/Netflix/dynomite/wiki
Ubuntu 20 in particular
I'm encountering the issue described here Netflix/dynomite#456
No idea how to proceed but I need to proceed
Any help?
“Nobody gets fired for picking redis enterprise”
It seems like antirez initiated redis clusters but since he "kind of" retired from Redis, I am not sure this would really move forward at least for the open source project.
Redis Entreprise seems to market itself as the way to scale Redis but it is so expensive and the pricing for Redis Enterprise Software is way too weird (per database shard) ...
It is also quite painful to maintain & deploy Redis Cluster day to day (node lost connection issue/ sharding issue after rejoin etc.).
The cheap way would be to use a proxy solution for Redis like Twitter/twemproxy but the project is quite dead, since the author left Twitter for starting an AI company that got bought by Etsy.
Plus, Twitter is now using Pelikan project and still struggling with reusing twemproxy.
Hopefully, Dynomite is reusing some part of Twemproxy and continuing the developing. So please continue the good work !
Redis Entreprise seems to market itself as the way to scale Redis but it is so expensive and the pricing for Redis Enterprise Software is way too weird (per database shard) ...
It is also quite painful to maintain & deploy Redis Cluster day to day (node lost connection issue/ sharding issue after rejoin etc.).
The cheap way would be to use a proxy solution for Redis like Twitter/twemproxy but the project is quite dead, since the author left Twitter for starting an AI company that got bought by Etsy.
Plus, Twitter is now using Pelikan project and still struggling with reusing twemproxy.
Hopefully, Dynomite is reusing some part of Twemproxy and continuing the developing. So please continue the good work !
FWIW, we still run Dynomite in production (the same as in the OSS repository) and it is in a pretty stable situation for us. We will continue using it as we see bugs we will fix them. In that sense, we have not stopped developing or releasing Dynomite, to some extent, it is just stable for our use cases. We would love for the community to pick up some steam and start developing features that they need. For more context, @smukil mentioned above that there are bandwidth issues in anwering the OSS questions.
@Marie-Diana yes Dynomite has used the baseline Twemproxy as a proxy but has extended it to become a separate project that does active-active.
Hi all. We are trying to setup dynomite to run in a docker-swarm. We notice that initial node restart few times because it feels like the DNS inside docker is some kind of slow to publish name resolution thus the dn_resolve at https://github.com/Netflix/dynomite/blob/0713e43618f74fe6fb21226287fcb3ec618bc2d3/src/dyn_util.c#L427 fails and we get the following log:
[2020-10-28 10:06:45.542] conf_handler:1228 conf: directive "dyn_seeds" has an invalid value
would it be possible to make this dns probe a bit more resilient by adding some kind of configurable retry logic?
Second question .. in those cases the dynomite crashes with a segfault with the following stacktrace
[2020-10-28 10:06:45.543] dn_stacktrace:286 [0] /lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fc039b52330]
??:0
[2020-10-28 10:06:45.548] dn_stacktrace:286 [1] src/dynomite(string_deinit+0x10) [0x439b19]
/dynomite/src/dyn_string.c:51
[2020-10-28 10:06:45.555] dn_stacktrace:286 [2] src/dynomite() [0x40ee03]
/dynomite/src/dyn_conf.c:142
[2020-10-28 10:06:45.562] dn_stacktrace:286 [3] src/dynomite() [0x40f4c9]
/dynomite/src/dyn_conf.c:291
[2020-10-28 10:06:45.568] dn_stacktrace:286 [4] src/dynomite(conf_destroy+0x4e) [0x413b2f]
/dynomite/src/dyn_conf.c:2165
[2020-10-28 10:06:45.575] dn_stacktrace:286 [5] src/dynomite(conf_create+0xde) [0x413ada]
/dynomite/src/dyn_conf.c:2155
[2020-10-28 10:06:45.584] dn_stacktrace:286 [6] src/dynomite() [0x416eea]
/dynomite/src/dyn_core.c:220
[2020-10-28 10:06:45.591] dn_stacktrace:286 [7] src/dynomite(core_start+0x22) [0x41702e]
/dynomite/src/dyn_core.c:253
[2020-10-28 10:06:45.597] dn_stacktrace:286 [8] src/dynomite() [0x441d75]
/dynomite/src/dynomite.c:557
[2020-10-28 10:06:45.604] dn_stacktrace:286 [9] src/dynomite(main+0x137) [0x441f78]
/dynomite/src/dynomite.c:624
[2020-10-28 10:06:45.611] dn_stacktrace:286 [10] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf5) [0x7fc039494f45]
??:0
[2020-10-28 10:06:45.615] dn_stacktrace:286 [11] src/dynomite() [0x40a039]
??:?
[2020-10-28 10:06:45.622] signal_handler:123 signal 11 (SIGSEGV) received, core dumping
??:0
[2020-10-28 10:06:45.548] dn_stacktrace:286 [1] src/dynomite(string_deinit+0x10) [0x439b19]
/dynomite/src/dyn_string.c:51
[2020-10-28 10:06:45.555] dn_stacktrace:286 [2] src/dynomite() [0x40ee03]
/dynomite/src/dyn_conf.c:142
[2020-10-28 10:06:45.562] dn_stacktrace:286 [3] src/dynomite() [0x40f4c9]
/dynomite/src/dyn_conf.c:291
[2020-10-28 10:06:45.568] dn_stacktrace:286 [4] src/dynomite(conf_destroy+0x4e) [0x413b2f]
/dynomite/src/dyn_conf.c:2165
[2020-10-28 10:06:45.575] dn_stacktrace:286 [5] src/dynomite(conf_create+0xde) [0x413ada]
/dynomite/src/dyn_conf.c:2155
[2020-10-28 10:06:45.584] dn_stacktrace:286 [6] src/dynomite() [0x416eea]
/dynomite/src/dyn_core.c:220
[2020-10-28 10:06:45.591] dn_stacktrace:286 [7] src/dynomite(core_start+0x22) [0x41702e]
/dynomite/src/dyn_core.c:253
[2020-10-28 10:06:45.597] dn_stacktrace:286 [8] src/dynomite() [0x441d75]
/dynomite/src/dynomite.c:557
[2020-10-28 10:06:45.604] dn_stacktrace:286 [9] src/dynomite(main+0x137) [0x441f78]
/dynomite/src/dynomite.c:624
[2020-10-28 10:06:45.611] dn_stacktrace:286 [10] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf5) [0x7fc039494f45]
??:0
[2020-10-28 10:06:45.615] dn_stacktrace:286 [11] src/dynomite() [0x40a039]
??:?
[2020-10-28 10:06:45.622] signal_handler:123 signal 11 (SIGSEGV) received, core dumping
looks like the deinit tries to remove a string which already turned to Null I believe the name of of the server configuration: string_deinit(&cs->name);
it would be nice to double check this and make the termination more graceful?
Hi guys, i am facing this issue when i try to start conductor server on dynomite db with redis backend
[ERROR] 2020-11-10 09:22:47,583 RedisMetadataDAO - refresh TaskDefs failed
com.netflix.dyno.connectionpool.exception.FatalConnectionException: FatalConnectionException: [host=Host [hostname=127.0.0.1, ipAddress=null, port=8102, rack: us-east-1e, datacenter: us-east-1, status: Up, hashtag=null, password=null], latency=0(0), attempts=1]redis.clients.jedis.exceptions.JedisConnectionException: Unexpected end of stream.
My config file is very simple
Database persistence model. Possible values are memory, redis, redis_cluster, redis_sentinel and dynomite.
If omitted, the persistence used is memory
#
memory : The data is stored in memory and lost when the server dies. Useful for testing or demo
redis : non-Dynomite based redis instance
redis_cluster: AWS Elasticache Redis (cluster mode enabled).See [http://docs.aws.amazon.com/AmazonElastiCache/latest/UserGuide/Clusters.Create.CON.RedisCluster.html]
redis_sentinel: Redis HA with Redis Sentinel. See [https://redis.io/topics/sentinel]
dynomite : Dynomite cluster. Use this for HA configuration.
db=dynomite
Dynomite Cluster details.
format is host:port:rack separated by semicolon
for AWS Elasticache Redis (cluster mode enabled) the format is configuration_endpoint:port:us-east-1e. The region in this case does not matter
workflow.dynomite.cluster.hosts=127.0.0.1:8102:us-east-1e
If you are running using dynomite, also add the following line to the property
to set the rack/availability zone of the conductor server to be same as dynomite cluster config
EC2_AVAILABILITY_ZONE=us-east-1e
EC2_REGION=us-east-1
Dynomite cluster name
workflow.dynomite.cluster.name=dyn_o_mite
Maximum connections to redis/dynomite
workflow.dynomite.connection.maxConnsPerHost=31
Namespace for the keys stored in Dynomite/Redis
workflow.namespace.prefix=conductor
Namespace prefix for the dyno queues
workflow.namespace.queue.prefix=conductor_queues
No. of threads allocated to dyno-queues (optional)
queues.dynomite.threads=10
Non-quorum port used to connect to local redis. Used by dyno-queues.
When using redis directly, set this to the same port as redis server.
For Dynomite, this is 22122 by default or the local redis-server port used by Dynomite.
queues.dynomite.nonQuorum.port=22122
I am asking here because i think it is an error in the communication between redis and dynomite, that i have started with docker following this guide http://www.dynomitedb.com/docs/dynomite/v0.5.8/quick-start/. Has anyone any idea? Thanks in advance.