Do you have the table definitions ?
Any example data?
Something is confused about a set / map / list type.
It's failing when replying the log, if you want to work around move the commit
log file out of the directory. There is a chance of data loss if this row
mutation is being replied on all n
Can your raise a ticket at https://issues.apache.org/jira/browse/CASSANDRA and
update the thread with the link?
Please include:
* nodetool status
* nodetool ring (so we have all the token assignments)
* The IP you started repair on
* As much log as you can share, if you can run DEBUG for the
o
Confirm if your write timeouts are client side socket time outs or the
TimedOutException from the server.
Typically write latency is related to GC problems, like you are seeing.
I'm unsure how much CPU resources each cassandra instance has. Is there one
node on a machine with 6 cores ?
How m
>
> cqlsh:Sessions> select * from "Items" where "mahoutItemid" =
> 610866442877251584;
>
> key| mahoutItemid
> +
> 687474703a2f2f6573706f7| 610866442877251584
>
> unsupported operand type(s) for /: 'NoneType' and 'float'
Can you
> "Bad Request: No indexed columns present in by-columns clause with Equal
> operator
> Perhaps you meant to use CQL 2? Try using the -2 option when starting cqlsh."
>
> My query is: select * from temp where min_update >10 limit 5;
You have to have at least one indexes column in the where clause
The broadcast_address can be set manually without using the
EC2MultiRegionSnitch. It's the address the node wants other nodes to talk to it
on
http://www.datastax.com/docs/1.2/configuration/node_configuration#broadcast-address
You may find it easier to run a VPN between the colo nodes and the
My general "I can haz heap space?" approach.
* determine total row count for the node from cfstats
* determine if wide (10's of MB) rows are in use
* determine total bloom filter space for the node from cfstats
* enable full GC logging as cassandra-env.sh
* determine tenured heap low point not lo
You may want to look at using virtual keyspaces:
http://hector-client.github.io/hector/build/html/content/virtual_keyspaces.html
And follow these tickets:
http://wiki.apache.org/cassandra/MultiTenant
-brian
On May 6, 2013, at 2:37 AM, Darren Smythe wrote:
> How many keyspaces can you reasonab
I was under the impression that it is multiple requests using a single
connectin PARALLEL not serial as they have request ids and the responses do as
well so you can send a request while a previous request has no response just
yet.
I think you do get a big speed advantage from the asynchronous
It seems that we did not have the JMX ports (1024+) opened in our firewall.
Once we opened ports 1024+ the hinted handoffs completed and it seems that the
cluster went back to normal.
Does that make sense?
Thanks,
Dan
This is what we saw in the logs after opening the ports:
INFO [HintedHandof
I had a 4 node cluster in my dev environment and due to resource
limitation, I had to remove two nodes. Nodetool status shows only two nodes
on both machines , but peers table on one machine still shows entries of
the nodes with a null rpc address. Thrift has no problem with it but new
Binary prot
What version of Cassandra are you using. If you're using 1.2.0 (or *were*
using 1.2.0 when the 2 nodes were removed), you might be seeing
https://issues.apache.org/jira/browse/CASSANDRA-5167.
> Or I have to delete the row in the table
That should work.
On Mon, May 6, 2013 at 4:22 PM, Shahryar S
Unfortunately I've just tried with a new cluster with RandomPartitioner and it
doesn't work better :
it may come from hadoop/pig modifications :
18:02:53|elia:hadoop cyril$ git diff --stat cassandra-1.1.5..cassandra-1.2.1 .
.../apache/cassandra/hadoop/BulkOutputFormat.java | 27 +--
.../apac
Just because you can batch queries or have the server process them out of
order doesn't make it fully "parellel". You're still using a single TCP
connection which is by definition a serial data stream. Basically, if you
send a bunch of queries which each return a large amount of data you've
effec
Hi Aaron,
The keyspace consistent of 3 column families for user management, see below.
I have dropped these tables multiple times since I'm testing a script to
automatically create the column families if they do not exists. I have
also been changing types, e.g. lock_tokens__ from MAP to
MAP.
On Sun, May 5, 2013 at 11:37 PM, Darren Smythe wrote:
> How many keyspaces can you reasonably have?
"Very Low Hundreds", though this relates more to CFs than Ks.
> If we have 1000 customers, so one KS per customer is 1000 keyspaces. Is that
> something C* can handle efficiently?
No.
> I guess
On Mon, May 6, 2013 at 6:20 AM, Dan Kogan wrote:
> It seems that we did not have the JMX ports (1024+) opened in our firewall.
> Once we opened ports 1024+ the hinted handoffs completed and it seems that
> the cluster went back to normal.
> Does that make sense?
No, JMX should not be required
On Sat, May 4, 2013 at 5:41 AM, Philippe wrote:
> After trying every possible combination of parameters, config and the rest,
> I ended up downgrading the new node from 1.1.11 to 1.1.2 to match the
> existing 3 nodes. And that solved the issue immediately : the schema was
> propagated and the node
On Sat, May 4, 2013 at 9:22 PM, Aiman Parvaiz wrote:
> We are using cassandra 1.1.0 and open-6-jdk
1.1.0 has significant issues, including non-working Hinted Handoff.
Also, OpenJDK is not officially supported.
Upgrade to 1.1.11 and Sun JDK.
=Rob
You have me thinking more. I wonder in practice if 3 sockets is any faster
than 1 socket when doing nio. If your buffer sizes were small, maybe that
would be the case. Usually the nic buffers are big so when the selector fires
it is reading from 3 buffers for 3 sockets or 1 buffer for one soc
Another option may be virtual column families with PlayOrm. We currently
do around 60,000 column families to store data from 60,000 different
sensors that keep feeding us information.
Dean
On 5/6/13 11:18 AM, "Robert Coli" wrote:
>On Sun, May 5, 2013 at 11:37 PM, Darren Smythe
>wrote:
>> How
I heard a rumor that index_interval is going away? What is the replacement for
this? (we have been having to play with this setting a lot lately as too big
and it gets slow yet too small and cassandra uses way too much RAM…we are still
trying to find the right balance with this setting).
Than
> De : aaron morton [mailto:aa...@thelastpickle.com]
> Envoyé : dimanche 28 avril 2013 22:54
> À : user@cassandra.apache.org
> Objet : Re: cost estimate about some Cassandra patchs
>
> > Does anyone know enough of the inner working of Cassandra to tell me how
> > much work is needed to patch Cassa
Correction, there was a typo in my original question, we are running cassandra
1.1.10
Thanks and sorry for the inconvenience.
On May 6, 2013, at 10:23 AM, Robert Coli wrote:
> including non-working Hinted Handoff
>From my experience, your NIC buffers generally aren't the problem (or at
least it's easy to tune them to fix). It's TCP. Simply put, your raw NIC
throughput > single TCP socket throughput on most modern hardware/OS
combinations. This is especially true as latency increases between the two
hosts
I think It will be better to open a issue in jira
Best regards
Shamim A.
> Unfortunately I've just tried with a new cluster with RandomPartitioner and
> it doesn't work better : > > it may come from hadoop/pig modifications : > >
> 18:02:53|elia:hadoop cyril$ git diff --
stat cassandra-1.1
Also have to keep in mind that it should be rare to only use a single
socket since you are usually making at least 1 connection per node in the
cluster (or local datacenter). There is also nothing enforcing that a
single client cannot open more than 1 connection to a node. In the end it
should come
Thanks. So then, Hinted Handoff should be sent over port 7000 (or 7001 with
SSL), correct?
-Original Message-
From: Robert Coli [mailto:rc...@eventbrite.com]
Sent: Monday, May 06, 2013 1:19 PM
To: user@cassandra.apache.org
Subject: Re: Node went down and came back up
On Mon, May 6, 201
On Mon, May 6, 2013 at 12:31 PM, Dan Kogan wrote:
> Thanks. So then, Hinted Handoff should be sent over port 7000 (or 7001 with
> SSL), correct?
Yes, hinted handoff goes over the "storage protocol" port, which is
shared with the "gossip" port, 7000/1.
=Rob
On Sat, May 4, 2013 at 9:22 PM, Aiman Parvaiz wrote:
>
> When starting this cluster we set
> > JVM_OPTS="$JVM_OPTS -Xss1000k"
>
>
>
Why did you increase the stack-size to 5.5 times greater than recommended?
Since each threads now uses 1000KB minimum just for the stack, a large
number of threads
> While reading we are planning to use a CL of Quorum. So, we are hoping we
> will not hit any consistency issues before repair is run.
There will be a chance of getting inconsistencies if less then QUORUM nodes
were involved in the load for each row. Assuming RF 3, if you have two adjacent
node
This is the closest I can find in Jira
https://issues.apache.org/jira/browse/CASSANDRA-4478
It's a pretty handy tool to have in your tool kit, specially when you start to
have over 1 billion rows per node.
A
-
Aaron Morton
Freelance Cassandra Consultant
New Zealand
@aaronmort
Thanks aaron..
--
Regards
Himanshu Joshi
On 05/06/2013 02:22 PM, aaron morton wrote:
"Bad Request: No indexed columns present in by-columns clause with Equal
operator
Perhaps you meant to use CQL 2? Try using the -2 option when starting cqlsh."
My query is: select * from temp where min_updat
33 matches
Mail list logo
