It might be an EC2 issue. We've been battling them all day and we noticed
Cassandra performance started to tank sometime earlier this week. I'm not
convinced the problem is with Cassandra, as we were seeing sub-100ms
response times prior to this and we haven't made any changes. We're also
running o
There's a lot more details that would be useful, but if you are on the
verge of OOMing and something actually running out, then that's
probably the culprit; when the JVM gets low on ram it will consume all
your CPU trying to GC enough to continue. (you mentioned seeing high
cpu on one core which t
On Fri, Feb 12, 2010 at 9:26 PM, Jonathan Ellis wrote:
>> And what is the reasoning behind lowering object count? Isn't it
>> whichever hits first which causes flushing? Or is there other
>> memory used when object count is higher?
>
> http://wiki.apache.org/cassandra/MemtableThresholds
These a
On Fri, Feb 12, 2010 at 6:08 PM, Anthony Molinaro
wrote:
> Also, is it okay to set MemtableSizeInMB in lower and restart?
Yes.
> And what is the reasoning behind lowering object count? Isn't it
> whichever hits first which causes flushing? Or is there other
> memory used when object count is h
right, and remembering that MemtableSizeInMB is just the size of the
values, you can estimate around 50% overhead on top of that.
On Fri, Feb 12, 2010 at 6:04 PM, Anthony Molinaro
wrote:
> Okay, is it better written as
>
> NumMemtables = 1 + 2 * AvailableProcessors + NumDataFileDirectory
>
> Thu
Hi,
I'm running a 5 node Cassandra cluster and am having a very tough time
getting reasonable performance from it. Many of the requests are failing
with TimeOutException. This is making it difficult to use Cassandra in a
production setting.
The cluster was running fine for a week or two (it was cr
Also, is it okay to set MemtableSizeInMB in lower and restart?
And what is the reasoning behind lowering object count? Isn't it
whichever hits first which causes flushing? Or is there other
memory used when object count is higher?
One other thing, is there other hidden memory (in cache, index, t
Okay, is it better written as
NumMemtables = 1 + 2 * AvailableProcessors + NumDataFileDirectory
Thus estimated maximum memory is
MemtableMemoryUsage = MemtableSizeInMB * NumMemtables
So in my case where I have 2 core machines with 3 datafile directories
MemtableMemoryUsage = 512 * (1 + 2 * 2 +
0.5 allows 1 + 2 * Runtime.getRuntime().availableProcessors()
Memtables + 1 per DataFileLocation to be waiting for flush before it
will block writes (or log replay) to give those time to flush out.
So, it sounds like you just need to lower your Memtable max
size/object count.
On Fri, Feb 12, 2010
0.5.0 final. I was able to get things going again by upping the memory
then lowering it after a successful restart, but I would like to know how
to minimize the chances of OOM via tuning.
-Anthony
On Thu, Feb 11, 2010 at 01:29:16PM -0600, Jonathan Ellis wrote:
> What version are you on these day
Also i have problem with StreamInitiateVerbHandler, the problem in
PendingFile.getTargetFile, namely difference in slashes on win and unix, so
i change PendingFile.java like this:
public PendingFile(String targetFile, long expectedBytes, String table)
{
targetFile_ = targetFile.rep
Care to include a stack trace? Those are useful when reporting problems.
On Fri, Feb 12, 2010 at 2:31 PM, ruslan usifov wrote:
> Yes
>
>
>
Yes
Does 64MB log an exception?
On Fri, Feb 12, 2010 at 10:45 AM, ruslan usifov wrote:
> My first node is ander win32, and there is problem in FileStreamTask.java
> here:
>
> public static final int CHUNK_SIZE = 64*1024*1024;
>
>
> On windows this buffer is very big so stream method of class FileStre
My first node is ander win32, and there is problem in FileStreamTask.java
here:
public static final int CHUNK_SIZE = 64*1024*1024;
On windows this buffer is very big so stream method of class FileStreamTask
always fail, i reduce buffer to 32MB, and this works in my case.
2010/2/12 ruslan usifo
If the processes are still active, could you please post a thread dump
of them? You can do this by sending the java process a kill -3
(assuming freebsd is similar enough to linux).
Gary
On Fri, Feb 12, 2010 at 04:47, ruslan usifov wrote:
> Hello
>
> I have two nodes. First i on only one node an
it is probably waiting for 192.168.0.37 to split off the right data
for it. send that command to 192.168.0.37.
-Jonathan
On Fri, Feb 12, 2010 at 4:47 AM, ruslan usifov wrote:
> Hello
>
> I have two nodes. First i on only one node and populate it with records.
> Then i start second node in boots
Hello
I have two nodes. First i on only one node and populate it with records.
Then i start second node in bootstrap mode. And it hung. I run ./nodetool
many times but it allways print
identical results (67108864/240394603) Nothing to grows. And CPU on
bootstrap node on 100%
freebsd# ./nodetool -
18 matches
Mail list logo
