?????? limit the sstable file size

2017-09-29 Thread Peng Xiao
Thanks Jeff for the quick reply. -- -- ??: "Jeff Jirsa";; : 2017??9??30??(??) 11:45 ??: "cassandra"; : Re: limit the sstable file size There's no way to limit file size in STCS. If you use LCS, it will default to 160MB (

Re: limit the sstable file size

2017-09-29 Thread Jeff Jirsa
There's no way to limit file size in STCS. If you use LCS, it will default to 160MB (except in cases where you have a very large partition - in those cases, the sstable will scale with your partition size, but you really shouldn't have partitions larger than 160MB) On Fri, Sep 29, 2017 at 8:41 P

limit the sstable file size

2017-09-29 Thread Peng Xiao
Dear All, Can we limit the sstable file size?as we have a huge cluster,the sstable file is too large for ETL to extract,Could you please advise? Thanks, Peng Xiao

Re: Nodetool repair -pr

2017-09-29 Thread Blake Eggleston
It will on 2.2 and higher, yes. Also, just want to point out that it would be worth it for you to compare how long incremental repairs take vs full repairs in your cluster. There are some problems (which are fixed in 4.0) that can cause significant overstreaming when using incremental repair.

Re: Issue with New Production Cluster

2017-09-29 Thread Michael Shuler
"Connection reset by peer" is almost certainly network issues. Same error: https://github.com/netty/netty/issues/5993 mtr - ping/trace tool to find possible flaky switch/router tcpdump and/or wireshark - tools to gather and observe network packets -- Michael On 09/29/2017 10:38 AM, Thakrar, Ja

Re: new question ;-) // RE: understanding batch atomicity

2017-09-29 Thread DuyHai Doan
We should probably replace "atomic" by "automatic retry" because it reflects exactly the actual guarantees On Fri, Sep 29, 2017 at 6:10 PM, Jon Haddad wrote: > The use of “atomic” for batches is misleading. Batches will eventually > complete, that doesn’t make them atomic. “All or nothing” is

Re: new question ;-) // RE: understanding batch atomicity

2017-09-29 Thread Jon Haddad
The use of “atomic” for batches is misleading. Batches will eventually complete, that doesn’t make them atomic. “All or nothing” is also incorrect, as you can read them in the middle and get “some parts of it”, and without a rollback it’s just “eventually all”. > On Sep 29, 2017, at 10:59 AM

Re: cassandra hardware requirements (STAT/SSD)

2017-09-29 Thread Jeff Jirsa
Cassandra was designed for spinning disks - commitlogs are (mostly) append-only, linear writes. sstables are written exactly once, again with linear writes. The index for finding the positions in sstables to start reads is cached in RAM (and when it's not cached, it's a linear read through a file)

Re: new question ;-) // RE: understanding batch atomicity

2017-09-29 Thread daemeon reiydelle
recall that a delete is actually a corner case of an update, as is an insert. As I read the snippet, you are updating multiple tables. The partition key is table specific, so two sets of update batches are handled here. We like to say that we don’t get to choose our parents, that they were given

new question ;-) // RE: understanding batch atomicity

2017-09-29 Thread DE VITO Dominique
Thanks DuyHai ! Does anyone know if BATCH provides atomicity for all mutations of a given partition key for a __single__ table ? Or if BATCH provides atomicity for all mutations of a given partition key for __ALL__ mutated tables into the BATCH ? That is, in case of : BEGIN BATCH Update table

Re: cassandra hardware requirements (STAT/SSD)

2017-09-29 Thread daemeon reiydelle
Note to the AWS poster, you have some limited understanding of how disks are presented to AWS compute nodes. As a result your post is not relevant, and misleading. When considering throughput, recall that disk IO is ideally parallel. While C* handles IO across multiple devices nicely, the unit of

Read-/ Write Latency - Cassandra 2.1 .15 vs 3.10

2017-09-29 Thread Mullachery, Anumod (Contractor)
Hi, We were running splunk queries to pull read / write latency. It's working fine in 2.1.15 , but not returning result from upgraded version 3.10. The bean used in the script is as shown below. Let me know, if any changes on the functionality on 2.1.15 vs 3.10 or it replaced to some other

Re: Issue with New Production Cluster

2017-09-29 Thread Jeff Jirsa
I don't know what logging is available driver side, I'd probably be writing a shell script to ping all three servers to see if I can prove there's a network problem outside of cassandra first. On Fri, Sep 29, 2017 at 8:19 AM, Jonathan Baynes < jonathan.bay...@tradeweb.com> wrote: > Thank you Jeff

Re: understanding batch atomicity

2017-09-29 Thread DuyHai Doan
All updates here means all mutations == INSERT/UPDATE or DELETE On Fri, Sep 29, 2017 at 5:07 PM, DE VITO Dominique < dominique.dev...@thalesgroup.com> wrote: > Hi, > > > > About BATCH, the Apache doc https://cassandra.apache.org/ > doc/latest/cql/dml.html?highlight=atomicity says : > > > > “*Th

understanding batch atomicity

2017-09-29 Thread DE VITO Dominique
Hi, About BATCH, the Apache doc https://cassandra.apache.org/doc/latest/cql/dml.html?highlight=atomicity says : "The BATCH statement group multiple modification statements (insertions/updates and deletions) into a single statement. It serves several purposes: ... All updates in a BATCH belongin

Re: Issue with New Production Cluster

2017-09-29 Thread Jeff Jirsa
The failure detector is seeing updates every 2.1-2.5 seconds, which it will ignore because it's over the 2 second default failure detector interval. If there's no load, there's no reason it shouldn't be seeing it far more frequently. If you're not seeing any signs of GC pauses (like GCInspector

RE: Issue with New Production Cluster

2017-09-29 Thread Jonathan Baynes
Hi Jeff, This is version 3.0.11. Being run on Oracle Red Hat Linux. If I retry immediately it fails, leave it for 20 minutes, like I have just now and retry it and it has worked. (?!?!) Ive checked all the logs (system and Debug ) and in the logs I have this: DEBUG [GossipStage:1] 2017-09-29 15

Re: Bootstraping a new node

2017-09-29 Thread Jeff Jirsa
Auto bootstrap doesn’t impact token assignment, it only impacts streaming - with it set to false a new node will instantly join the cluster on startup without streaming any data. -- Jeff Jirsa > On Sep 29, 2017, at 4:53 AM, Jean Carlo wrote: > > Thx @Jacob, I was reading that article, I don

Re: Issue with New Production Cluster

2017-09-29 Thread Jeff Jirsa
What version? If you retry immediately, does it reconnect? Anything in the logs? What you describe is atypical - timeouts on queries can (and will) happen occasionally under load, but timeout on connect is atypical. Any sign of networking issues/slowness/dns problems/etc? On Fri, Sep 29, 2017 at

Issue with New Production Cluster

2017-09-29 Thread Jonathan Baynes
Hi Community, I have a 6 node ring, covering 2 DC's, the ring isn't being used yet and we are just in the connectivity and testing phase. So the boxes are NOT under any load. I've gone to connect to CQLSH this afternoon and I've had this returned: cqlsh xx.xxx.xxx.xx -u cassandra -p cassandra C

How Can I get started with Using Cassandra and Netbeans- Please help

2017-09-29 Thread Lutaya Shafiq Holmes
How Can I get started with Using Cassandra and Netbeans- Please help -- Lutaaya Shafiq Web: www.ronzag.com | i...@ronzag.com Mobile: +256702772721 | +256783564130 Twitter: @lutayashafiq Skype: lutaya5 Blog: lutayashafiq.com http://www.fourcornersalliancegroup.com/?a=shafiqholmes "The most beautif

Re: cassandra hardware requirements (STAT/SSD)

2017-09-29 Thread Lutaya Shafiq Holmes
Please try and USE AWS amazon web services on aws.amazon.com On 9/29/17, Peng Xiao <2535...@qq.com> wrote: > Hi there, > we are struggling on hardware selection,we all know that ssd is good,and > Datastax suggests us to use ssd,as Cassandra is a CPU bound db,we are > considering to use sata disk,

Proposal for deprecating/removing the read_repair_chance/dclocal_read_repair_chance table options

2017-09-29 Thread Sylvain Lebresne
We are considering deprecating and them ultimately removing the 2 table options: 'read_repair_chance' and 'dclocal_read_repair_chance'. The rational and much more details are on CASSANDRA-13910 (https://issues.apache.org/jira/browse/CASSANDRA-13910), so I won't repeat it here. The goal of this ema

Re: Bootstraping a new node

2017-09-29 Thread Jean Carlo
Thx @Jacob, I was reading that article, I dont understand how we can add a node in the cluster using auto_bootstrap equal to false and not having a maldistribution of tokens. I know that cassandra use random token ranges if it doesn't have tokens associated. So it doesn't compute the new tokens ra

Re: Bootstraping a new node

2017-09-29 Thread Jacob Shadix
Are you building out a new DC? This is a good article about all things re. bootstrap - http://thelastpickle.com/blog/2017/05/23/auto-bootstrapping-part1.html Refer to the auto_bootstrap: false and see if that meets your use case. Pay CLOSE attention to the caveats. Bootstrapping w/o streaming ris

Bootstraping a new node

2017-09-29 Thread Jean Carlo
Hello cassandra community When bootstraping a new node, there is a way to say to cassandra from which DC make the streaming ? Or there is a way to just join the machine to the cluster just calculating the new tokens ( I am using vnodes) and not making any streaming to later make a rebuild from k