I think he actually meant *increase*, for this reason "For small T, a
random choice of initial tokens will in most cases give a poor distribution
of data. The larger T is, the closer to uniform the distribution will be,
with increasing probability."
Alain
2013/6/11 Theo Hultberg
> thanks, tha
thanks, that makes sense, but I assume in your last sentence you mean
decrease it for large clusters, not increase it?
T#
On Mon, Jun 10, 2013 at 11:02 PM, Richard Low wrote:
> Hi Theo,
>
> The number (let's call it T and the number of nodes N) 256 was chosen to
> give good load balancing for
You can refer to this conversation here
http://comments.gmane.org/gmane.comp.db.cassandra.user/27366
Manoj
On Tue, Jun 11, 2013 at 10:01 AM, Tanya Malik wrote:
> If I understand the C* architecture correctly, in order to increase write
> speed, I only need to put the commit log on SSDs.
>
> Whe
Hi,
Does C* support something like co-processor functionality/triggers to run
client-supplied code in the address space of the server?
In the older versions it was possible, but, in C* 1.2 it is a global
configuration so you won't be able to configure it per CF basis.
Manoj
On Tue, Jun 11, 2013 at 10:32 AM, Tanya Malik wrote:
> Is it possible in C* 1.2 to configure column families to be flushed
> individually?
>
> So, if I hav
Is it possible in C* 1.2 to configure column families to be flushed
individually?
So, if I have 3 CFs, and one of them gets full to 64 MB and the others are
only at 5 MB, will only the first full CF get flushed to a SStable?
Also, is it possible to configure different sized memtables for differen
If I understand the C* architecture correctly, in order to increase write
speed, I only need to put the commit log on SSDs.
When the memtable gets flushed to the SStable file later on, that can go to
traditional spinning disks, since that happens much later after the write
successful has already b
On Mon, Jun 10, 2013 at 3:13 PM, Emalayan Vairavanathan
wrote:
> I suspect that nodetool cleanup is IO intensive. So running nodetool cleanup
> concurrently on the entire cluster may have a significantly impact the IO
> performance of applications.
cleanup is a specific kind of compaction, and as
-- Forwarded message --
From: "Mark Lewandowski"
Date: Jun 8, 2013 8:03 AM
Subject: Cassandra (1.2.5) + Pig (0.11.1) Errors with large column families
To:
Cc:
> I'm cur.rently trying to get Cassandra (1.2.5) and Pig (0.11.1) to play
nice together. I'm running a basic script:
>
>
Thank you Edward.
I suspect that nodetool cleanup is IO intensive. So running nodetool cleanup
concurrently on the entire cluster may have a significantly impact the IO
performance of applications.
Apart from this, do you see any other implications on running the nodetool
cleanup concurrently
You eventually should run cleanup to remove data no longer needed on the
node. However it does not need to be run quickly after a join. You can run
it when you get around to it. I would run it on a few nodes at a time until
they are all cleaned up.
On Mon, Jun 10, 2013 at 5:00 PM, Emalayan Vairav
Hi Theo,
The number (let's call it T and the number of nodes N) 256 was chosen to
give good load balancing for random token assignments for most cluster
sizes. For small T, a random choice of initial tokens will in most cases
give a poor distribution of data. The larger T is, the closer to unifo
Hi All,
Datastax manual suggests that during a Cassandra cluster expansion, an
administrator has to run nodetool cleanup on each of the previously existing
Cassandra nodes to remove the keys that are no longer belonging to those nodes.
Further the manual says that thenodetool cleanup task shou
Thanks for all the advice I really appreciate it.
1.) Seeing as how I only had a single node cluster previously, if I just
`nodetool move 0` the original node that should be an easy fix then?
2.) Is my data not marked as deleted? If I use sstableloader to restream the
data am I just reorganizi
This is true unless you have ran cleanup. If you have run 'nodetool
cleanup' then there is data loss.
On Mon, Jun 10, 2013 at 1:40 PM, Tyler Hobbs wrote:
>
> On Mon, Jun 10, 2013 at 11:39 AM, Nimi Wariboko Jr <
> nimiwaribo...@gmail.com> wrote:
>
>> How can I recover that data? Can I assume the
I'm not sure I follow what you mean, or if I've misunderstood what
Cassandra is telling me. Each node has 256 vnodes (or tokens, as the
prefered name seems to be). When I run `nodetool status` each node is
reported as having 256 vnodes, regardless of how many nodes are in the
cluster. A single node
On Mon, Jun 10, 2013 at 11:39 AM, Nimi Wariboko Jr
wrote:
> How can I recover that data? Can I assume they are still in the sstables?
Yes
> Would doing a sstable2json then reading and reinserting be an optimal
> solution?
You can just use the sstableloader utility to load them directly. It
To recover it would be to dump everything then re-insert everything.
Another option would be to return all nodes to whatever tokens they were
before the switches, since the old data is still there.
Either way both recovery options are long,painful, and a good amount of
manual steps. I would not w
I thought it was interesting that when I tried to subscribe to the developers
list my e-mail was bounced for being spam. But the user subscription got
through just fine.
Could this be happening on the unsubscribe? I've seen "your message is spam"
messages bounced by the recipient for looking
How can I recover that data? Can I assume they are still in the sstables? Would
doing a sstable2json then reading and reinserting be an optimal solution?
On Monday, June 10, 2013 at 9:18 AM, Tyler Hobbs wrote:
>
> On Sun, Jun 9, 2013 at 3:19 PM, Nimi Wariboko Jr (mailto:nimiwaribo...@gmail.co
I unsubscribed a while ago and then resubscribed. It took about four
unsubscribe attempts before it actually worked.
From: "Fatih P."
Reply-To:
Date: Mon, 10 Jun 2013 17:46:30 +0300
To:
Subject: Re: Unsubscribe?
i tried the same and receiving mails.
On Mon, Jun 10, 2013 at 5:34 PM, Luke
Although there is a ticket open that WILL start removing data
automatically. Which I keep saying is a bad idea.
On Mon, Jun 10, 2013 at 12:18 PM, Tyler Hobbs wrote:
>
> On Sun, Jun 9, 2013 at 3:19 PM, Nimi Wariboko Jr
> wrote:
>
>> If I had to do a repair after upping the RF, than that is pro
reading changelog for eclipse kepler (4.3) and BIRT has support for
creating reports from Cassandra
On Sun, Jun 9, 2013 at 3:19 PM, Nimi Wariboko Jr wrote:
> If I had to do a repair after upping the RF, than that is probably what
> caused the data loss. Wish I had been more careful.
>
> I'm guessing the data is irrevocably lost, I didn't make any any snapshots.
>
> Would it be possible to figure
You sent an email to user-unsubscr...@cassandra.apache.org from the
email addressed used, and it didn't unsubscribe you? Did you get the
'are you sure' email? Did you check your spam folder?
see
http://cassandra.apache.org/
http://hadonejob.com/img/70907344.jpg
On 06/10/2013 10:46 AM, Fati
There are n vnodes regardless of the size of the physical cluster.
Regards
Milind
On Jun 10, 2013 7:48 AM, "Theo Hultberg" wrote:
> Hi,
>
> The default number of vnodes is 256, is there any significance in this
> number? Since Cassandra's vnodes don't work like for example Riak's, where
> there i
Hi,
The default number of vnodes is 256, is there any significance in this
number? Since Cassandra's vnodes don't work like for example Riak's, where
there is a fixed number of vnodes distributed evenly over the nodes, why so
many? Even with a moderately sized cluster you get thousands of slices.
i tried the same and receiving mails.
On Mon, Jun 10, 2013 at 5:34 PM, Luke Hospadaruk wrote:
> Hi,
> I hate to be a clod, but I'd really like to unsubscribe from this list.
> I've tried every permutation I can think of to do it "the right way", and
> all of the styles in the help message. If
Hi,
I hate to be a clod, but I'd really like to unsubscribe from this list. I've
tried every permutation I can think of to do it "the right way", and all of the
styles in the help message. If there's a moderator reading this could you
please take me off the list?
Thanks,
Luke
Hi,
If one increases the replication factor of a keyspace and then do a
repair, how will this affect the performance of the affected nodes?
Could we risk the nodes being (more or less) unresponsive while repair
is going on? The nodes I am speaking of contains ~100gb of data.
Also, some of the key
Sounds good, feel free to text me when you are there.
Later,
Dean
On 6/7/13 3:38 PM, "Faraaz Sareshwala" wrote:
>I'll be attending and will try and meet up with you :). I see your posts
>often
>on this list -- would love to pick your brain and learn more about what
>you are
>using cassandra fo
Yep, *MetaData stuff is what I was looking for thanks for the pointer.
Joe
On Sun, Jun 9, 2013 at 8:00 PM, Sylvain Lebresne wrote:
> DESCRIBE is not a CQL3 query but a cqlsh specific command, so it won't
> work outside of cqlsh. However, at least if you are using the datastax java
> driver (I d
32 matches
Mail list logo
