Re: Multiple SSD disks per sever? Ideal config?

2014-11-06 Thread Kevin Burton
I have seen people do this but I can’t find documentation for it, and specifically how well it optimizes IO. Does it write blocks to both disks? How is IO parallelized. Too many questions to list them all. On Thu, Nov 6, 2014 at 3:27 PM, Chris Lohfink wrote: > If optimizing for IO, use Cassan

Re: tuning concurrent_reads param

2014-11-06 Thread Jimmy Lin
I see, thanks for explaining what that means. If we are using SSD, then reordering/merging has less impact than traditional mechanical hard disk, so using SSD drive probably can deal with increased concurrent_read better. (?)

Re: tuning concurrent_reads param

2014-11-06 Thread Bryan Talbot
On Wed, Nov 5, 2014 at 11:00 PM, Jimmy Lin wrote: > Sorry I have late follow up question > > In the Cassandra.yaml file the concurrent_read section has the following > comment: > > What does it mean by " the operations to enqueue low enough in the stack > that the OS and drives can reorder t

Re: Multiple SSD disks per sever? Ideal config?

2014-11-06 Thread Chris Lohfink
If optimizing for IO, use Cassandra's JBOD configuration (list each disk under data directories in cassandra.yaml). It would put sstables on the disk thats least used. If want to optimize for disk space, I'd go with RAID0. Will probably want to tune concurrent reader/writers, stream throughput (

Multiple SSD disks per sever? Ideal config?

2014-11-06 Thread Kevin Burton
I’m curious what people are doing with multiple SSDs per server. I think there are two main paths: - RAID 0 them… the problem here is that RAID0 is not a panacea and the drives may or may not see better IO throughput. - use N cassandra instances per box (or containers) and have one C* node acces

Re: C* on Fusion IO

2014-11-06 Thread Kevin Burton
On Thu, Nov 6, 2014 at 2:10 PM, Christopher Brodt wrote: > Yep. The "trouble" with FIOs is that they almost completely remove your > disk throughput problems, so then you're constrained by CPU. Concurrent > compactors and concurrent writes are two params that come to mind but there > are likely o

Re: C* on Fusion IO

2014-11-06 Thread Kevin Burton
This is definitely a first world problem.. having databases that are CPU bound :-P On Thu, Nov 6, 2014 at 1:05 PM, jeeyoung kim wrote: > I've been running with FIOs and we've been CPU bound most of the time. But > I'm not using native transport yet, and is hoping that it would make things > fast

Re: C* on Fusion IO

2014-11-06 Thread Christopher Brodt
Yep. The "trouble" with FIOs is that they almost completely remove your disk throughput problems, so then you're constrained by CPU. Concurrent compactors and concurrent writes are two params that come to mind but there are likely others. @kevin. I hear you. 5TB is sort of a maximum that DataStax

Re: C* on Fusion IO

2014-11-06 Thread jeeyoung kim
I've been running with FIOs and we've been CPU bound most of the time. But I'm not using native transport yet, and is hoping that it would make things faster. On Thu, Nov 6, 2014 at 12:54 PM, Christopher Brodt wrote: > You should get pretty great performance with those FusionIO cards. One > thin

Re: C* on Fusion IO

2014-11-06 Thread Kevin Burton
This was one of my biggest issues too. We were expecting to be at 5-10 nodes to start with and then 20-40 nodes in 60-90 days. But this means we can run all of our database on 1 box :-P … but realistically two. Which means if one box goes offline then I’m at 50% capacity. That and I don’t even

Re: C* on Fusion IO

2014-11-06 Thread Christopher Brodt
You should get pretty great performance with those FusionIO cards. One thing I watch out for whenever scaling Cassandra vertically is compaction times, which probably won't matter here. However, you have to take into account that you lose some resiliency to failures with less nodes. On Thu, Nov 6,

Re: C* on Fusion IO

2014-11-06 Thread Russ Bradberry
I've heard of people running dense nodes (8+ TB) using fusion I/O, but with 10GBe connections. I mean why buy a Ferrari and never leave first gear? As far as saturating the network goes, I guess that all depends on your workload, and how often you need to repair. Sent from my iPhone > On Nov 6

C* on Fusion IO

2014-11-06 Thread Kevin Burton
We’re looking at switching data centers and they’re offering pretty aggressive pricing on boxes with fusion IO cards. 2x 1.2TB Fusion IO 128GB RAM 20 cores. now.. this isn’t the typical cassandra box. Most people are running multiple nodes to scale out vs scale vertically. But these boxes are p

Re: read after write inconsistent even on a one node cluster

2014-11-06 Thread Jonathan Haddad
For cqlengine we do quite a bit of write then read to ensure data was written correctly, across 1.2, 2.0, and 2.1. For what it's worth, I've never seen this issue come up. On a single node, Cassandra only acks the write after it's been written into the memtable. So, you'd expect to see the most

Re: read after write inconsistent even on a one node cluster

2014-11-06 Thread Robert Coli
On Thu, Nov 6, 2014 at 6:14 AM, Brian Tarbox wrote: > We write values to our keyspaces and then immediately read the values back > (in our Cucumber tests). About 20% of the time we get the old value.if > we wait 1 second and redo the query (within the same java method) we get > the new value

authentication with cassandra-stress 2.1

2014-11-06 Thread James Derieg
Is there a way to authenticate to cassandra using the new cassandra-stress tool released with cassandra 2.1? It appears as if the '-un' (username) and '-pw' (password) switches have been removed from the tool. In the 2.0 version, this is the command I would run: 'cassandra-stress -D nodesfile

Re: read after write inconsistent even on a one node cluster

2014-11-06 Thread Brian Tarbox
Thanks. Right now its just for testing but in general we can't guard against multiple users ending up the one writes and then one reads. It would be one thing if the read just got old data but we're seeing it return wrong data...i.e. data that doesn't correspond to any particular version of the

Re: read after write inconsistent even on a one node cluster

2014-11-06 Thread Eric Stevens
If this is just for doing tests to make sure you get back the data you expect, I would recommend looking some sort of eventually construct in your testing. We use Specs2 as our testing framework, and our write-then-read tests look something like this: someDAO.write(someObject) eventually { s

read after write inconsistent even on a one node cluster

2014-11-06 Thread Brian Tarbox
We're doing development on a single node cluster (and yes of course we're not really deploying that way), and we're getting inconsistent behavior on reads after writes. We write values to our keyspaces and then immediately read the values back (in our Cucumber tests). About 20% of the time we get

Re: Storing files in Cassandra with Spring Data / Astyanax

2014-11-06 Thread DuyHai Doan
You'd better off asking on the Spring Data Cassandra mailing list. I think that very few people not to say nobody tried integrating Astyanax with Spring Data Cassandra... Le 6 nov. 2014 08:17, "Wim Deblauwe" a écrit : > Hi, > > We are building an application where we install it on-premise, usual

Re: Counter column impossible to delete and re-insert

2014-11-06 Thread DuyHai Doan
Hello Clément This is a known anti-pattern. You should never re-use a deleted counter column otherwise there will be unpredictable result for the counter value. Le 6 nov. 2014 08:45, "Clément Fumey" a écrit : > Hi, > > I have a table with counter column . When I insert (update) a row, delete > i