Since the 0.7 upgrade, I've been going through and scrubbing all of
our sstables on 0.7.4. Some of the tables have completely unordered
keys, and the scrub fails to work on those tables. In those cases, I
export the sstable via sstable2json, and reimport it with
json2sstable.
Tonight I've ran into
When the node starts it reads the stored token information from the
LocationInfo CF in the System KS.
It looks like the log message "is now part of the cluster" is only logged when
an endpoint is added to a nodes view of the ring via gossip It is not logged
when the endpoint is added during st
Which version are you using ?
It looks like using 0.7X (and prob 0.6) versions you can just shutdown the node
and bring it back up with the new IP and It Just Works
https://issues.apache.org/jira/browse/CASSANDRA-872
I've not done it before, anyone else ?
Aaron
On 23 Mar 2011, at 07:53, Cas
FWIW: For whatever reason jna memlockall does not work for us. jna call is
successful but cassandra process swaps anyway.
see: http://www.mail-archive.com/user@cassandra.apache.org/msg11235.html
We disabled swap entirely.
On Mar 22, 2011, at 8:56 PM, Chris Goffinet wrote:
> The easiest way to
Hello,
what kind of bug is it?
If I do nodetool host1 ring, the output is:
Address Status State LoadOwnsToken
141784319550391026443072753096570088105
1.174 Up Normal 4.14 GB 16.67% 0
1.173 Down Normal 4.07 GB 16.67%
283568639100782052886145
Hello
After having updgraded from 0.7.0 to 0.7.4, cassandra does not start
anymore.
I have the following errot stack:
INFO 11:47:04,112 Finished reading
/var/lib/cassandra/commitlog/CommitLog-1296815168287.log
ERROR 11:47:04,114 Exception encountered during startup.
java.lang.NullPointerExceptio
Did you drop a keyspace not too long before the upgrade, or did you
mess up with your system tables
in the process ?
--
Sylvain
On Wed, Mar 23, 2011 at 11:47 AM, Sébastien Druon wrote:
> Hello
> After having updgraded from 0.7.0 to 0.7.4, cassandra does not start
> anymore.
> I have the followin
- Original Message -
From: buddhasystem
To: cassandra-u...@incubator.apache.org
Sent: Tue Mar 22 21:49:39 2011
Subject: Re: 0.7.2 choking on a 5 MB column
I see. I'm doing something even more drastic then, because I'm only inserting
one row in this case, and just use cf.insert(), witho
2011/3/23 aaron morton
> Which version are you using ?
>
> It looks like using 0.7X (and prob 0.6) versions you can just shutdown the
> node and bring it back up with the new IP and It Just Works
> https://issues.apache.org/jira/browse/CASSANDRA-872
>
>
>
So to replace one machine on another, the
Then that shouldn't happen. Could you open a ticket on JIRA for this ?
If you could set the DEBUG log level for log4j and include the parts
relevant to the
error in the ticket description, that could help.
--
Sylvain
On Wed, Mar 23, 2011 at 2:16 PM, Sébastien Druon wrote:
> Hi,
> I just installe
Sure! :)
[]'s
FernandoVM
On Tue, Mar 22, 2011 at 5:07 PM, aaron morton wrote:
> Sounds interesting, please let the community know your findings.
>
> Aaron
>
> On 23 Mar 2011, at 01:31, FernandoVM wrote:
>
>> Hi,
>>
>>> contrib/py_stress is the easiest way to shake out any issues with your
>>> in
I'm going through the process of specing out the hardware for a
Cassandra cluster. The relevant specs:
- Support 460 operations/sec (50/50 read/write workload). Row size
ranges from 4 to 8K.
- Support 29 million objects for the first year
- Support 365 GB storage for the first year, based on Cassa
Hello
Sometimes i seen in gc log follow message:
2011-03-23T14:40:56.049+0300: 14897.104: [GC 14897.104: [ParNew (promotion
failed)
Desired survivor size 41943040 bytes, new threshold 2 (max 2)
- age 1:5573024 bytes,5573024 total
- age 2:5064608 bytes, 10637632 total
: 672577K->
On Tue, Mar 22, 2011 at 6:49 PM, buddhasystem wrote:
> I see. I'm doing something even more drastic then, because I'm only inserting
> one row in this case, and just use cf.insert(), without batch mutator. It
> didn't occur to me that was a bad idea.
Everything in a given RPC has to complete befo
Hi maki-san
I am so sorry this was my mistake.
I expected when I set data at one node, the data should be copied to the
other node
at the same keyspace and same column family by replication.
This replication was working.
I just made a mistake to get data by wrong Key.(First character of key
wa
I think it is due to fragmentation in old gen, due to which survivor area
cannot be moved to old gen. 300MB data size of memtable looks high for 3G
heap. I learned that in memory overhead of memtable can be as high as 10x of
memtable data size in memory. So either increase the heap or reduce the
me
2011/3/23 Narendra Sharma
> I think it is due to fragmentation in old gen, due to which survivor area
> cannot be moved to old gen. 300MB data size of memtable looks high for 3G
> heap. I learned that in memory overhead of memtable can be as high as 10x of
> memtable data size in memory. So eithe
I understand that. The overhead could be as high as 10x of memtable data
size. So overall the overhead for 16CF collectively in your case could be
300*10 = 3G.
Thanks,
Naren
On Wed, Mar 23, 2011 at 11:18 AM, ruslan usifov wrote:
>
>
> 2011/3/23 Narendra Sharma
>
>> I think it is due to fragment
What process are you using to confirm the data was replicated to another server
? And what makes you say the data is not replicated ? I think your
understanding of the replication may be a little off, you rarely read from one
node. Have a look at
http://thelastpickle.com/2011/02/07/Introduction
First thing is check the logs on host 1. Check the view of the ring from all
other nodes in the cluster, do they think nodes 2 and 3 are also down? Then
confirm all nodes have the same config for listen port and all nodes can telnet
to the listen port for the other nodes.
I'm guessing the inse
I saw this once when my servers ran out of file descriptors. This caused
totally weird problems.
Make sure all nodes in the cluster are listening on the gossip port (7000 by
default).
Also check out
http://www.datastax.com/docs/0.7/troubleshooting/index#view-of-ring-differs-between-some-nodesor
7000 and 9160 are accessible. Don't think I need other ports for basic
setup , right ?
If anyone coud get 'nodetool repair' working with this patch (across
regions), let me know. It may be I am doing something wrong.
On Wed, Mar 23, 2011 at 1:08 AM, Milind Parikh wrote:
> @aj
> are you sure that
I'm going to take a stab at a hypothesis:
Sunday: I drain and decommission 2.3.4.193 *but* I forget to run node cleanup
on the rest of the nodes. The ring looks clean but I did not see "Annoucing
that ..." in the logs.
Tuesday: the ghost node reappears on the ring for all nodes.
Could this be c
Hi Aaron,
I'm using 7.3 (upgrading to 7.4). Basically I needed to swap the IPs of two
of my nodes. Unfortunately, after I did so, neither was accessible in the
original ring anymore. I saw the warning message in the logs about the
token being reassigned. Maybe I didn't give it enough time, but
2011/3/23 Narendra Sharma
> I understand that. The overhead could be as high as 10x of memtable data
> size. So overall the overhead for 16CF collectively in your case could be
> 300*10 = 3G.
>
>
> And how about G1 GC, it must prevent memory fragmentation. but some post on
this email, told that i
After an upgrade from 0.7.3 to 0.7.4, we're seeing the following on
several data files:
ERROR [main] 2011-03-23 18:58:33,137 ColumnFamilyStore.java (line 235)
Corrupt sstable
/mnt/services/cassandra/var/data/0.7.4/data/Helium/dp_idx-f-4844=[Index.db,
Statistics.db, Data.db, Filter.db]; skipped
jav
Hi Aaron
Thank you so much for your reply and advice.
I watched your presentation. it"s helpful for me.
Anyway I did as following with 2 node servers.(53;1st node,54;2nd node)
I started the following write/read.php program by through Thrift at 53
server.
$users->insert($Key, array('Movie'
Does anybody know if it's possible to find out what node a specific key/row
lives on?
We have a 30 node cluster and I'm curious how much faster it'll be to read
data directly from the node that stores the data.
We're using random partitioner, by the way.
*Sameer Farooqui
*Accenture Technology L
On Wed, Mar 23, 2011 at 2:23 PM, Alexis Lê-Quôc wrote:
> I'm going to take a stab at a hypothesis:
> Sunday: I drain and decommission 2.3.4.193 *but* I forget to run node
> cleanup on the rest of the nodes. The ring looks clean but I did not see
> "Annoucing that ..." in the logs.
>
> Tuesday: t
On Wed, Mar 23, 2011 at 4:59 PM, Erik Onnen wrote:
> After an upgrade from 0.7.3 to 0.7.4, we're seeing the following on
> several data files:
>
> ERROR [main] 2011-03-23 18:58:33,137 ColumnFamilyStore.java (line 235)
> Corrupt sstable
> /mnt/services/cassandra/var/data/0.7.4/data/Helium/dp_idx-f
Can any of the dev's jump in here ? Whats best practice ?
Aaron
On 24 Mar 2011, at 08:32, Casey Deccio wrote:
> Hi Aaron,
>
> I'm using 7.3 (upgrading to 7.4). Basically I needed to swap the IPs of two
> of my nodes. Unfortunately, after I did so, neither was accessible in the
> original r
Each row is stored on RF nodes, and your read will be sent to CL number of
nodes. Messages only take a single hop from the coordinator to each node the
read is performed on, so the networking overhead varies with the number of
nodes involved in the request. There are man factors other than netw
I haven't used G1. I remember someone shared his experience in detail on G1.
The bottom line is you need to test it for your deployment and based on test
and results conclude if it will work for you. I believe for a small heap G1
will do well.
-Naren
On Wed, Mar 23, 2011 at 1:47 PM, ruslan usifo
Thanks, so is it the "[Index.db, Statistics.db, Data.db, Filter.db];
skipped" that indicates it's in Statistics? Basically I need a way to
know if the same is true of all the other tables showing this issue.
-erik
No problems with read performance, just curious about what kind of overhead
was being added b/c we're doing read tests.
If it's easy to figure out where the row is stored, I'd be interested in
trying it. If not, don't worry about it.
- Sameer
On Wed, Mar 23, 2011 at 4:31 PM, aaron morton wrote:
On Wed, Mar 23, 2011 at 4:31 PM, aaron morton wrote:
> There are features available to determine which nodes holds replicas for a
> particular key. AFAIK they are not intended for use by clients.
Specifically :
http://wiki.apache.org/cassandra/JmxInterface#org.apache.cassandra.service.StorageSer
The logic to find the node is not complicated. You compute the MD5 hash of
the key. Create sorted list of tokens assigned to the nodes in the ring.
Find the first token greater than the hash. This is the first node. Next in
the list is the replica, which depends on the RF. Now this is simple becau
It's been about 7 months now but at the time G1 would regularly
segfault for me under load on Linux x64. I'd advise extra precautions
in testing and make sure you test with representative load.
On Wed, Mar 23, 2011 at 6:52 PM, Erik Onnen wrote:
> Thanks, so is it the "[Index.db, Statistics.db, Data.db, Filter.db];
> skipped" that indicates it's in Statistics? Basically I need a way to
> know if the same is true of all the other tables showing this issue.
It's the at
org.apache.cassand
My nodetool repair does not hang. That's why I'm curious.
/***
sent from my android...please pardon occasional typos as I respond @ the
speed of thought
/
On Mar 23, 2011 2:54 PM, "A J" wrote:
7000 and 9160 are accessible. Don't think I need other por
If you are getting the "Cluster schema does not agree" then you have a sick
cluster and replication will not be working properly.
Open the cassandra-cli client and type "describe cluster;" you should see a
single schema version such as...
[default@unknown] describe cluster;
Cluster Informatio
It really does depend on what your workload is like, and in the end will
involve a certain amount of fudge factor.
http://wiki.apache.org/cassandra/CassandraHardware provides some guidance.
http://wiki.apache.org/cassandra/MemtableThresholds can be used to get a rough
idea of the memory requ
Hey guys,
I have a one node system(with replication factor of 1) running cassandra.
The machine has two disks. One is used as the commitlog and the other as
cassandra's data directory. The node just had gotten unresponsive and had to
be hard rebooted.
After restart, cassandra started off fine. But
Hi,
This problem occurs when the clause has multi expression and a expression with
operator other than EQ.
Is anyone meet the same problem?
I trace the code, and seen this at ColumnFamilyStore.satisfies() method:
int v = data.getComparator().compare(column.value(),
expression.value)
Looks like this https://issues.apache.org/jira/browse/CASSANDRA-2347
From this discussion
http://www.mail-archive.com/user@cassandra.apache.org/msg11291.html
Aaron
On 24 Mar 2011, at 17:17, Wangpei (Peter) wrote:
> Hi,
>
> This problem occurs when the clause has multi expression and a expre
This looks like a bug
(https://issues.apache.org/jira/browse/CASSANDRA-2376), but not one
that would cause a crash. Actual process death is only caused by (a)
running out of memory or (2) JVM bugs.
On Wed, Mar 23, 2011 at 9:17 PM, Sanjeev Kulkarni wrote:
> Hey guys,
> I have a one node system(wit
We support changing IPs to _new_ IPs but I'd be surprised if it can
handle changing it to the IP of another node. I'd try using a 3rd
"temporary" IP if I wanted to swap two.
On Wed, Mar 23, 2011 at 6:18 PM, aaron morton wrote:
> Can any of the dev's jump in here ? Whats best practice ?
> Aaron
>
hi,
i'm trying to run cassandra 0.7 on my windows machine,
and i don't seem to be able to go beyond the warning message
C:\Program Files\Apache Software
Foundation\apache-cassandra-0.7.4\bin>cassandra
-f
Starting Cassandra Server
log4j:WARN No appenders could be found for logger
(org.apache.cassa
48 matches
Mail list logo
