ueries to key spaces that are only replicated in a
single region and I use LOCAL_SERIAL CL
would 100 CAS queries per second that in the normal case do not conflict
(== work in different partition keys) be sort of 'ok'?
Or should it rather be in the range of 10/s?
Jan
[1] https://www
reads all quorum CLs will yield more requests sent by the
coordinator to other nodes and hence *QUORUM reads definitely increase
cluster load. (And of course response time of the coordinator, too).
Correct?
Jan
-
To unsubscribe, e
misses an event that only later pops up during repair.
What that happens, I'd like to re-process the log (my processing is
idempotent, so it can just go again).
This is why I was looking for a way to learn that a repair has actually
repaired something.
Jan
On Mon, May 29, 2017 at
Hi,
is it possible to extract from repair logs the writetime of the writes
that needed to be repaired?
I have some processes I would like to re-trigger from a time point if
repair found problems.
Is that useful? Possible?
Jan
SSTables bloat.
Makes sense?
Jan
On Fri, May 26, 2017 at 7:41 AM Max C wrote:
In my case, we're using Cassandra to store QA test data — so the
pattern
is that we may do a bunch of updates within a few minutes / hours,
and then
the data will essentially be read-only for the rest of
Hi Jayesh,
On 25 May 2017, at 18:31, Thakrar, Jayesh wrote:
Hi Jan,
I would suggest looking at using Zookeeper for such a usecase.
thanks - yes, it is an alternative.
Out of curiosity: since both, Zk and C* implement Paxos to enable such
kind of thing, why do you think Zookeeper would be
to mostly happen in
the memtable resulting in only occasional manifestation in SSTables.
Is that assumption correct and if so, what config parameters should I
tweak to keep the memtable from being flushed for longer periods of
ti
m on a 2.2.8 cluster).
Jan
-
To unsubscribe, e-mail: user-unsubscr...@cassandra.apache.org
For additional commands, e-mail: user-h...@cassandra.apache.org
It seems that sstablesplit cant handle the "new" filename pattern
anymore (acutally running 2.2.8 on those nodes).
Any hints or other suggestions to split those sstables or get rid of them?
Thanks in advance,
Jan
--
Centers and a
RF of 3.
Has anyone encountered this problem and if yes what steps have you
taken to solve it
Thanks,
Charu
--
Jan Kesten, mailto:j.kes...@enercast.de
Tel.: +49 561/4739664-0 FAX: -9 Mobil: +49 160 / 90 98 41 68
enercast GmbH Universitätsplatz 12 D-34127 Kassel HRB15471
http
Hi,
could you post the output of nodetool cfstats for the table?
Cheers,
Jan
Am 16.02.2017 um 17:00 schrieb Selvam Raman:
> I am not getting count as result. Where i keep on getting n number of
> results below.
>
> Read 100 live rows and 1423 tombstone cells for query S
Hi,
do you got a result finally?
Those messages are simply warnings telling you that c* had to read many
tombstones while processing your query - rows that are deleted but not
garbage collected/compacted. This warning gives you some explanation why
things might be much slower than expected be
often (about 2500 times a minute) and I was
wondering if this is just "ok" or if there is something misusing paged
results for requests fetching a single record and we should have a look
at it. Maybe paging results could be a performance issue?
Thanks for any hints,
Jan
seperate disks when using
spindles.
Third, have you monitored iostats and cpustats while running your tests?
Cheers,
Jan
Am 08.02.2017 um 16:39 schrieb Branislav Janosik -T (bjanosik - AAP3 INC
at Cisco):
Hi all,
I have a cluster of three nodes and would like to ask some questions
about the
Hi,
I am looking for a driver for the Rust language. I found some projects
which seem quite abandoned.
Can someone point me to the driver that makes the most sense to look at
or help working on?
Cheers,
Jan
https://www.azul.com/products/zing/order-zing/
At least a list price for zing I found there: 3k$ per year.
- Ursprüngliche Nachricht -
Von: "Work"
Gesendet: 26.11.2016 07:53
An: "user@cassandra.apache.org"
Betreff: Re: Java GC pauses, reality check
I'm not affiliated with them, I've
Hi,
can you check the size of your data directories on that machine to verify in
comparison to the others?
Have a look for snapshot directories which could still be there from a former
table or keyspace.
Regards,
Jan
Am 26. Oktober 2016 06:53:03 MESZ, schrieb Harikrishnan A :
>Hello,
>
oo deep into this - maybe 2.1.16 or 2.2.8 - as chances
are really good your problems will be gone after that.
Regards.
Jan
the job - but if you do not need
data from one of the keyspaces at all just drop and recreate it (but
look into your data directory if there are snapshots left). Prevent this
in future have a close look at heap consumption and maybe give it more
memory.
HTH,
Jan
map should lead to a delete.
Is that correct?
Jan
96)
at org.apache.cassandra.config.Schema.(Schema.java:50)
at org.apache.cassandra.tools.nodetool.Cleanup.execute(Cleanup.java:45)
at org.apache.cassandra.tools.NodeTool$NodeToolCmd.run(NodeTool.java:248)
at org.apache.cassandra.tools.NodeTool.main(NodeTool.java:162)
Can anyone help me?
Best re
HI Folks;
I am trying to have one of my DSE 4.7 C* nodes also function as a Solr node
within the cluster.
I have followed the docs in vain :
https://docs.datastax.com/en/datastax_enterprise/4.0/datastax_enterprise/srch/srchInstall.html
Any pointers would help.
Thanks
Jan
- you can repeat last two steps and use sstableload only on tables with
mtime > timestamp to add the differencens to cluster1
- shutdown cluster2 when done
Of course, data written by old clients to cluster2 wont be available in
cluster1 until loading that data into it.
Just my 2 cents :)
Jan
been implemented.
Recommend reading up this blog article:
http://www.datastax.com/dev/blog/whats-coming-to-cassandra-in-3-0-improved-hint-storage-and-delivery
hope this helps
Jan/
On Thu, 4/21/16, Jens Rantil wrote:
Subject: Re: When are hints
used, the data
will stream from the decommissioned node. If removetoken is used, the data will
stream from the remaining replicas.
Hope this helps
Jan/
On Thu, 4/21/16, Anubhav Kale wrote:
Subject: RE: Problem Replacing a Dead Node
To: "
'nodetool repair' on cluster 1
d. de-commission cluster2.
You are ready to use cluster 1 [with both keyspaces within it]
Hope this helps
Jan
On Thu, 4/21/16, Arlington Albertson wrote:
Subject: Combining two clusters/keyspaces i
Jens;
I am unsure that you need to enable Replication & also use the sstable loader.
You could load the data into the new DC and susbsequently alter the keyspace to
replicate from the older DC.
Cheers
Jan
On Thu, 4/21/16, Jens Rantil w
Hi,
you should check the "snapshot" directories on your nodes - it is very
likely there are some old ones from failed operations taking up some space.
Am 15.04.2016 um 01:28 schrieb kavya:
Hi,
We are running a 6 node cassandra 2.2.4 cluster and we are seeing a
spike in the disk Load as per
also take into account to store the keys (hashes) in a
seperate table per day / hour or something like that, so you can quickly
get all keys for a time range. A query without the partition key may be
very slow.
Jan
Am 11.04.2016 um 23:43 schrieb Robert Wille:
I have a need to be able to use t
as the gps satellites are flying
atom clocks :)
Just my 2 cents,
Jan
Von meinem iPhone gesendet
> Am 31.03.2016 um 03:07 schrieb Mukil Kesavan :
>
> Hi,
>
> We run a 3 server cassandra cluster that is initially NTP synced to a single
> physical server over LAN. This server d
), column1, column2)
)
Cheers,
Jan
them again online, much less
files to copy now. After that I shutdown the node and my last rsync now has to
copy only a few files which is quite fast and so the downtime for that node is
within minutes.
Jan
Von meinem iPhone gesendet
> Am 18.02.2016 um 22:12 schrieb Branton Davis :
>
&
needs more understanding and planning.
Just as a hint and offtopic: I saw people using cassandra as application glue
for interprocess communication where every app server started a node (for
communication, sessions and as queue and so on). If that is eventually a use
case - have a look at hazelcast.
SizeTieredCompaction you can end up with very huge sstables as I do
(>250gb each). In the worst case you could possibly need twice the space - a
reason why I set up my monitoring for disk to 45% usage.
Just my 2 cents.
Jan
Von meinem iPhone gesendet
> Am 13.02.2016 um 08:48 schrieb Branton Davis :
>
B
E C F
Any input would be much appreciated.
thanks,Jan
Keep in mind that compaction in LCS can only run 1 compaction per level.
Even if it wants to run more compactions in L0 it might be blocked
because it is already running a compaction in L0.
BR
Jan
On 01/16/2016 01:26 AM, Sebastian Estevez wrote:
LCS is IO ontensive but CPU is also relevant
HI Folks
could you please point me to the 2015 Cassandra summit held in California. I
do see the ones posted for the 2014 & 2013 conferences.
ThanksJan
cassandra.yaml to remove the additional datadir
- shutdown the node
- rsync again (just for the case, a new sstable got written while the
first one was running)
- restart
HTH
Jan
Am 14.01.2016 um 08:38 schrieb Rahul Ramesh:
> One update. I cleared the snapshot using nodetool clearsnapshot comm
consuming your space.
Jan
Von meinem iPhone gesendet
> Am 14.01.2016 um 07:25 schrieb Rahul Ramesh :
>
> Thanks for your suggestion.
>
> Compaction was happening on one of the large tables. The disk space did not
> decrease much after the compaction. So I ran an external c
crub.
Any suggestions what is causing this?
Thanks in advance,
Jan
to do the work?
So instead of saying 'for this query, LOCAL_SERIAL is enough for me' this would
be like saying 'I want XYZ to happen exactly once, per data center'. - All
services would try to do XYZ, but only one instance *per datacenter* will
actually become the leader and succeed.
Makes sense?
Jan
David ;
bring down all the nodes with the exception of the 'seed' node.Now bring up the
10th node. Run 'nodetool status' and wait until this 10th node is UP. Bring
up the rest of the nodes after that. Run 'nodetool status' again and check
that all the nodes are UP.
Alternatively;decommiss
nodetool cfstats
would be your best bet. Sum all the column families info., within a keyspace
to get to the number you are looking for.
Jan/
On Wednesday, July 1, 2015 9:05 AM, graham sanderson
wrote:
If you are pushing metric data to graphite, there is
The request would return with the latest data.
The read request would fire against node 1 and node 3. The coordinator would
get answers from both and would merge the answers and return the latest.
Then read repair might run to update node 3.
QUORUM does not take into consideration whether an an
Marcin ;
are all your nodes within the same Region ? If not in the same region,
what is the Snitch type that you are using ?
Jan/
On Thursday, April 2, 2015 3:28 AM, Michal Michalski
wrote:
Hey Marcin,
Are they actually going up and down repeatedly (flapping) or just
moving average, I have to deal with data all
over the place. I can't currently think of anything but performing
aggregateByKey causing a shuffle every time.
Anyone have experience with combining time series chunking and computation
on all / many time series at once? Any advice?
Cheers,
Frens Jan
HI Jatin;
besides enabling Tracing, is there any other way to get the task done ? (to
log the client ID for every operation)Please share with the community the
solution, so that we could collectively learn from your experience.
cheersJan/
On Friday, February 20, 2015 12:48 PM, Jatin
Paul Nickerson;
curious, did you get a solution to your problem ?
Regards,Jan/
On Tuesday, February 10, 2015 5:48 PM, Flavien Charlon
wrote:
I already experienced the same problem (hundreds of thousands of SSTables)
with Cassandra 2.1.2. It seems to appear when running an
querying from the second table ?
Unfortunately, I have more questions that answers; however despite the
sacrilege of using super-columns (lol), there has got to be a logical answer to
the Performance problem you are having. Hopefully we could dig in and
find an answer .
Jan/
HI Batranut;
In both errors you described above the files seem to be missing while
compaction is running. Without knowing what else is going on your system, I
would presume that this error occurs on this single node only and not your
entire cluster.
Some guesses:a) You may have a disk corrupt
Benyi ;
have you considered using the TTL option in case your columns are meant to be
deleted after a predetermined amount of time ? Its probably the easiest way to
get the task accomplished.
cheersJan
On Friday, February 27, 2015 10:38 AM, Benyi Wang
wrote:
In C* 2.1.2, is there
HI Rahul;
your question: Can we see active queries on cassandra cluster. Is there any
tool?
Answer: nodetool tpstats & nodetool cfsstats The nodetool tpstats
command provides statistics about the number of active, pending, and completed
tasks for each stage of Cassandra operations by th
Ian;
to respond to your specific question:
You could pipe the output of your repair into a file and subsequently determine
the time taken. example: nodetool repair -dc DC1
[2014-07-24 21:59:55,326] Nothing to repair for keyspace 'system'
[2014-07-24 21:59:55,617] Starting repair command #2, re
in Luxembourg to DSE 4.6.1 h) conduct a
'nodetool repair -parallel' again i) Upgrade to OpsCenter 5.1
Best of luck, hope this helps.
Jan/
On Wednesday, March 18, 2015 1:01 PM, Robert Coli
wrote:
On Wed, Mar 18, 2015 at 9:05 AM, David CHARBONNIER
wrote:
-
David;
all the packaged installations use the /var/lib/cassandra directory. Could you
check your yaml config files and see if you are using this default directory
for backups
May want to change it to a location with more disk space.
hope this helpsJan/
On Monday, March 16, 2015 2:5
77 to -594461978511041000 were
included. In a case which yielded much more partition keys, the entire
token range did seem to be queried.
To reiterate my initial questions: is this behavior to be expected? Am I
doing something wrong? Is there a workaround?
Best regards,
Frens Jan
On 4 March
March 2015 at 18:10, DuyHai Doan wrote:
> First idea to eliminate any issue with regards to staled data: issue the
> same count query with RF=QUORUM and check whether there are still
> inconsistencies
>
> On Tue, Mar 10, 2015 at 9:13 AM, Rumph, Frens Jan
> wrote:
>
&g
You could set up an Alert for Node down within OpsCenter. OpsCenter also
offers you the option to send an email to a paging system with reminders.
Jan/
On Sunday, March 8, 2015 6:10 AM, Vasileios Vlachos
wrote:
We use Nagios for monitoring, and we call the following through
HI Folks;
We are planning to deploy a Multi region C* Cluster with nodes on both US
coasts. Need some advice :
a) As I do not have Public IP address access, is there an alternative way to
deploy EC2MultiRegion snitch using Private IP addresses ? b) Has anyone used
EC2_Snitch with nodes
x27;s strange it only happen in this node but this type of message does not
shown in other node log file at the same time...
Jason
On Thu, Mar 5, 2015 at 4:26 AM, Jan wrote:
HI Jason;
Whats in the log files at the moment jstat shows 100%. What is the activity on
the cluster & the node at the
Hello Jaydeep;
Run cassandra-stress with R/W options enabled for about the same time and
check if you have dropped packets. It would eliminate the client as the source
of the error & also give you a replicable tool to base subsequent tests/
findings.
Jan/
On Thursday, March 5,
HI Jason;
Whats in the log files at the moment jstat shows 100%. What is the activity on
the cluster & the node at the specific point in time (reads/ writes/ joins etc)
Jan/
On Wednesday, March 4, 2015 5:59 AM, Jason Wee wrote:
Hi, our cassandra node using java 7 update 72 an
HI Jaydeep;
- look at the i/o on all three nodes
- Increase the write_request_timeout_in_ms: 1
- check the time-outs if any on the client inserting the Writes
- check the Network for dropped/lost packets
hope this helpsJan/
On Wednesday, March 4, 2015
Heap
The leveled compaction issue is not addressed by this. hope this helps
Jan/
On Wednesday, March 4, 2015 8:41 AM, Roni Balthazar
wrote:
Hi there,
We are running C* 2.1.3 cluster with 2 DataCenters: DC1: 30 Servers /
DC2 - 10 Servers.
DC1 servers have 32GB of RAM and 10GB of HEAP. DC2
2.1.2 on 64bit fedora 21 with Oracle
Java 1.8.0_31.
Thanks in advance,
Frens Jan
hey have only 3 TB drives. I made a screenshot.
https://www.dropbox.com/s/0qhbpm1znwd07rj/strange_sizes.png?dl=0
Did this occur somewhere else? Maybe it is totally unrelated to 2.1.3
upgrade.
Thanks for any pointers,
Jan
ing a repartition is way to expensive as I just want more
partitions for parallelism, not reshuffle ...
Thanks in advance!
Frens Jan
Hi Batranut,
apart from the other suggestions - do you have ntp running on all your
cluster nodes and are times in sync?
Jan
rite implements full usage of
https://github.com/datastax/spark-cassandra-connector and brings much of
it's goodness to PySpark!
Hope that some of you are able to put this to good use. And feedback, pull
requests, etc. are more than welcome!
Best regards,
Frens Jan
/CASSANDRA-8839
Jan
HI Asit;
The Partition key is only a part of the performance. Recommend reading this
article: Advanced Time Series with Cassandra
| |
| | | | | | | |
| Advanced Time Series with CassandraDataStax - Software, support, and training
for Apache Cassandra |
| |
| View on www.datast
HI Gabriel;
I don't think Apache Cassandra supports in-memory keyspaces. However Datastax
Enterprise does support it.
Quoting from Datastax: DataStax Enterprise includes the in-memory option for
storing data to and accessing data from memory exclusively. No disk I/O occurs.
Consider using the
Colin;
Ceph is a block based storage architecture based on RADOS. It comes with its
own replication & rebalancing along with a map of the storage layer.
Some more details & similarities: a)Ceph stores a client’s data as objects
within storage pools. (think of C* partitions)b) Using the
a Consultant Pythian - Love your data
rolo@pythian | Twitter: cjrolo | Linkedin:
linkedin.com/in/carlosjuzarteroloTel: 1649www.pythian.com
On Sat, Jan 31, 2015 at 2:49 AM, Asit KAUSHIK
wrote:
Hi all,
We are testing our logging application on 3 node cluster each system is virtual
machine wi
Saurabh;
a) How exactly are the three nodes hosted. b) Can you take down node 2 and
create the keyspace from node 1c) Can you take down node 1 and create the
keyspace from node2d) Do the nodes see each other with 'nodetool status'
cheersJan/
C* Architect
On Saturday, January 31
HI Michal;
The consistency level defaults to ONE for all write and read operations.
However consistency level is also set for the keyspace.
Could it be possible that your queries are spanning multiple keyspaces which
bear different levels of consistency ?
cheersJan
C* Architect
On Frida
going in the right direction.
Hope this helpsJan/
On Thursday, January 29, 2015 5:01 PM, Jan wrote:
Thanks Rob;
here is what I am looking for :
java -jar /home/user/jmxterm-1.0-alpha-4-uber.jar 10.30.41.52:7199 -O
org.apache.cassandra.internal:type=FlushWriter -A CurrentlyBlocke
Mbean: org.apache.cassandra.request
Attribute: org.apache.cassandra.request:type=ReadStage
Hope this helpsJan/
On Thursday, January 29, 2015 9:13 AM, Batranut Bogdan
wrote:
Hello,
Is there a metric that will show how many reads per second C* serves? Read
requests shows ho
:
On Thu, Jan 29, 2015 at 3:27 PM, Jan wrote:
I am trying to use JMXterm, a command line based tool to script & monitor C*
cluster.
Would anyone on this forum know the exact syntax to connect to Cassandra Domain
using JMXterm ?
Here's an example from an old JIRA at my shop :
1.
HI Folks;
I am trying to use JMXterm, a command line based tool to script & monitor C*
cluster. Would anyone on this forum know the exact syntax to connect to
Cassandra Domain using JMXterm ?Please give me an example.
I do 'not' intend to use OpsCenter or any other UI based tool.
thanksJan
cable test result, I recommend the following: a) Keep the
'data' expectation to a point in time which is a known quanta. b) Load some
data into your cluster & take a snapshot. Reload this snapshot before every
Test for consistent results.
hope this helps.
Jan/C* Architect
|
| |
| View on techblog.netflix.com | Preview by Yahoo |
| |
| |
May want to chase Jeff Magnuson & check if the solution is open sourced. Pl.
report back to this forum if you get an answer to the problem.
hope this helps. Jan
C* Architect
On Monday, January 26, 2015 11:25 AM, Ro
could only include the Primary & clustered
keys and it should be fine.
You identify the new row via : Primary & clustered keys.
Errata: You could add Longitude & Latitude too to the model to add a level
of detail especially since its widely prevalent for weather station data.
hope this
Running a 'nodetool repair' will 'not' bring the node down.
Your question: does a nodetool repair make the server stop serving requests, or
does it just use a lot of ressources but still serves request
Answer: NO, the server will not stop serving requests. It will use
some resource
ing your read performance, check nodetool cfstats
and nodetool cfhistograms .
On Thu, Jan 15, 2015 at 2:11 AM, Roland Etzenhammer
mailto:r.etzenham...@t-online.de>> wrote:
Hi,
I'm testing around with cassandra fair a bit, using 2.1.2 which I
know has some major issu
compaction you will have one active SSTable5 which is newly written and
consumes space. The snapshot-linked ones are still there, still
consuming their space. Only when this snapshot is cleared you get your
disk space back.
HTH,
Jan
at java.lang.Thread.run(Thread.java:744) [na:1.7.0_51]
BR
Jan
Hi,
even if recovery like a dead node would work - backup and restore (like
my way with an usb docking station) will be much faster and produce less
IO and CPU impact on your cluster.
Keep that in Mind :-)
Cheers,
Jan
Am 22.12.2014 um 10:58 schrieb Or Sher:
Great. replace_address works
- everything should be fine ;-)
Of course you will need a replication factor > 1 for this to work ;-)
Just my 2 cents,
Jan
rsync the full contents there,
Am 18.12.2014 um 16:17 schrieb Or Sher:
Hi all,
We have a situation where some of our nodes have smaller disks and we
would like to al
`dirname $0`/../../lib/*.jar; do
-CLASSPATH=$CLASSPATH:$jar
done
+elif [ -r "$CASSANDRA_INCLUDE" ]; then
+. "$CASSANDRA_INCLUDE"
fi
# Use JAVA_HOME if set, otherwise look for java in PATH
---SNIP---
Worked for me on both tools.
Jan
Hi Jens,
maybe you should have a look at mutagen for cassandra:
https://github.com/toddfast/mutagen-cassandra
It is a litte quiet around this for some months, but maybe still worth it.
Cheers,
Jan
Am 25.11.2014 um 10:22 schrieb Jens Rantil:
Hi,
Anyone who is using, or could recommend, a
@cassandra.apache.org
Subject: Re: high context switches
On Fri, Nov 21, 2014 at 1:21 AM, Jan Karlsson
mailto:jan.karls...@ericsson.com>> wrote:
Nothing really wrong with that however I would like to understand why these
numbers are so high. Have others noticed this behavior? How much context
switch
Hello,
We are running a 3 node cluster with RF=3 and 5 clients in a test environment.
The C* settings are mostly default. We noticed quite high context switching
during our tests. On 100 000 000 keys/partitions we averaged around 260 000 cs
(with a max of 530 000).
We were running 12 000~ tran
Hi all,
thanks again for the comments.
I have created an (improved?) design, this time using dedicated consumers per
shard and time-based row expire, hence without immediate deletes.
https://github.com/algermissen/cassandra-ruby-sharded-workers
As before, comments are welcome.
Jan
On 06 Oct
that the goal primarily is to keep the rows ‘short’
enough to achieve a tombstones read performance impact that one can live with
in a given use case.
Is that understanding wrong?
Jan
to do one insert to put the job in the queue and another insert to mark the
> job as done or in process
> or whatever. This would also give you the benefit of being able to replay the
> state of the queue.
Thanks, I’ll try that, too.
Jan
>
>
> On Mon, Oct 6, 2014 at 12:5
quite understood what Netfix is doing in terms of
coordination - but since performance isn’t our concern, CAS should do fine, I
guess(?)
Thanks again,
Jan
>
> ---
> Chris Lohfink
>
>
> On Oct 5, 2014, at 6:03 PM, Jan Algermissen
> wrote:
>
>> Hi,
>>
thoughts on
the design path I took.
Jan
[1] https://github.com/Netflix/astyanax/wiki/Message-Queue
r during automated tests
If you delete with T1 and insert with T1 the delete wins, which was the reason
in our case.
You might want to test this with client provided timestamps and make sure the
insert has a T_insert > T_delete
Jan
>
> Is it a bug on Cassandra or on Datastax driver?
> Any suggestions?
>
> Tks
/magro/play2-scala-cassandra-sample
The actual mapping from Java to Scala futures for the async case is in
https://github.com/magro/play2-scala-cassandra-sample/blob/master/app/models/Utils.scala
HTH,
Jan
> Thanks
1 - 100 of 185 matches
Mail list logo
