Hi everyone,
Due to a crash, we lost the system_auth keyspace. It was not important at that
time because this was a closed system, not communicating outside the structure
so we used AllowAllAuthorizer.
Now it is changing and we would like to recover that keyspace (re-create it
from scratch).
U
Good time of day everyone;
Does Cassandra have a "recommended uptime"? I.e., does regular Cassandra
node reboots help anything? Are periodical node reboots recommended for
general system stability?
Best regards, Vsevolod.
Hello,
Those tables are empty by default, except system_auth.roles table which
contains one entry(by default): the cassandra user/role.
CassandraRoleManager creates it on startup if it does not exist.
Those tables will be populated once you start adding new users/roles
and use authorization.
Wha
Hi Horia,
Thanks for your reply :-)
As the keyspace was corrupt, Cassandra didn’t boot… So as we were in a hurry
and not needed that keyspace, we dropped the files and put AllowAllAuthorizer.
Now even the roles table is empty.
When I enable PasswordAuthenticator, I’m not able to authenticate, c
Try executing in cqlsh:
insert into system_auth.roles (role , can_login , is_superuser ,
salted_hash ) VALUES ( 'cassandra', True, True,
'$2a$10$qQIh9pXDu0JNA7vQd7KRcO7VXryjbTu8XBQTC.tXcTpJHliH6S1aW');
The above hash encrypted value is for the default password cassandra
After doing this you shou
Not anything that I'm aware of. Cassandra can run at months/years without
rebooting.
It is better to monitor your nodes and if you find anything abnormal a restart
can help.
--
SIMON FONTANA OSCARSSON
Software Developer
Ericsson
Ölandsgatan 1
37133 Karlskrona, Sweden
simon.fontana.oscars...@er
It’s likely that if you have server stability issues its because of data model
or compaction strategy configurations which lead to out of memory issues or
massive GC pauses. Rebooting wouldn’t solve those issues.
--
Rahul Singh
rahul.si...@anant.us
Anant Corporation
On Jul 17, 2018, 7:28 AM -04
What’s the goal, Abdul? Is it for security reasons or for organizational
reasons. You could try prefixing / suffixing the keyspace names if its for
organizational reasons (For now) if you don’t want to do the manual management
of mounts as Anthony suggested .
--
Rahul Singh
rahul.si...@anant.us
Have you considered looking into reaper project — could save you time in
figuring out your own strategy.
https://github.com/thelastpickle/cassandra-reaper
Otherwise you can always do a round robin of cron jobs per node once a week…
Your repair cycle should repair all servers within a window les
You can make new clusters or you can isolate with datacenters that don’t have a
keyspace replicated.
On Jul 16, 2018, 10:41 AM -0400, Durity, Sean R ,
wrote:
> In most cases, we separate clusters by application. This does help with
> isolating problems. A bad query in one application won’t affec
Hi,
Thanks I tried that, made a node tool repair system_auth and I get a new error
now :
Connection error: ('Unable to connect to any servers', {'127.0.0.1':
AuthenticationFailed('Failed to authenticate to 127.0.0.1: Error from server:
code=0100 [Bad credentials] message="Provided username cas
@rahul.xavier.si...@gmail.com, @simon.fontana.oscars...@ericsson.com,
Thank you for answers.
I've got a really heavy data model, but no GC problems.
Moreover, gclog shows very efficient GC process and hardly any GC pauses.
So, the question remains open: how general is the practice of
semi-perio
What are general community and/or your personal experience viewpoints on
cassandra node RAM amount vs data stored per node question?
Thank you very much.
Best regards,
Vsevolod.
I haven’t restarted a node (service) unless there are changes made to config,
and for restarting the server its incredibly rare. In fact since Production has
been up I think ive restrted the servers twice (both for DR occurences) in a
Year.
There is very little need to reboot the servers and un
What Cassandra version do you use?
On tis, 2018-07-17 at 14:23 +0200, Thomas Lété wrote:
> Hi,
>
> Thanks I tried that, made a node tool repair system_auth and I get a
> new error now :
>
> Connection error: ('Unable to connect to any servers', {'127.0.0.1':
> AuthenticationFailed('Failed to aut
The latest : 3.11.2, the same as yours :(
> Le 17 juil. 2018 à 14:36, Horia Mocioi a écrit :
>
> What Cassandra version do you use?
>
> On tis, 2018-07-17 at 14:23 +0200, Thomas Lété wrote:
>> Hi,
>>
>> Thanks I tried that, made a node tool repair system_auth and I get a
>> new error now :
>>
We do not have any scheduled, periodic node restarts. I have been working on
Cassandra across many versions, and I have not seen a case where periodic
restarts would solve any problem that I saw.
There are certainly times when a node needs a restart – but those are because
of specific reasons.
The requirement is that we plan to have a new keyspace which will hist PII
data and we wanted to see if we can have encrypted ssd then can we find the
new keyspace with PII can be binded to it and no chanhes would required to
current keyspaces.
On Tuesday, July 17, 2018, Rahul Singh wrote:
> Wha
Could you try the following steps?
Stop Cassandra.
Change authenticator in yaml to PasswordAuthenticator if not already done.
Remove data directory with `rm -rf data/system_auth/roles-*`
Start Cassandra.
Login with `cqlsh -u cassandra -p cassandra`
Works for me.
--
SIMON FONTANA OSCARSSON
Softw
Ok I tried that, nothing better (I already tried dropping the entire
system_auth folder that way, same result)
When I open the log, I found nothing about « Password » and when I search for «
roles », I only find that :
DEBUG [main] 2018-07-17 15:37:39,420 CompactionStrategyManager.java:380 -
R
Could you also send the output of "select * from system_auth.roles"?
(you will need to change authenticator to AllowAllAuthenticator and
authorizer to AllowAllAuthorizer)
On tis, 2018-07-17 at 15:43 +0200, Thomas Lété wrote:
> Ok I tried that, nothing better (I already tried dropping the entire
>
It’s empty...
> Le 17 juil. 2018 à 15:59, Horia Mocioi a écrit :
>
> Could you also send the output of "select * from system_auth.roles"?
> (you will need to change authenticator to AllowAllAuthenticator and
> authorizer to AllowAllAuthorizer)
>
> On tis, 2018-07-17 at 15:43 +0200, Thomas Lété
This is very strange behavior if Cassandra won't recreate the cassandra user
when you delete the folder.
So just to make sure, you are stopping Cassandra on all nodes and deleting the
data directory?
--
SIMON FONTANA OSCARSSON
Software Developer
Ericsson
Ölandsgatan 1
37133 Karlskrona, Sweden
Should not need regular reboots for Cassandra’s sake, but you should be
regularly patching your servers
--
Jeff Jirsa
> On Jul 17, 2018, at 2:09 AM, Vsevolod Filaretov wrote:
>
> Good time of day everyone;
>
> Does Cassandra have a "recommended uptime"? I.e., does regular Cassandra node
>
Yes I did that multiple time, always following the same procedure : stop
Cassandra, on all nodes, remove data, update config then restart nodes one by
one…
I really don’t understand when I could have done wrong...
> Le 17 juil. 2018 à 16:15, Simon Fontana Oscarsson
> a écrit :
>
> This is ve
I usually don’t want to put more than 1.0-1.5 TB ( at the most ) per node. It
makes streaming slow beyond my patience and keeps the repair / compaction
processes lean. Memory depends on how much you plan to keep in memory in terms
of key / row cache. For my uses, no less than 64GB if not more ~
@Rahul Singh thank you for the answer!
What is your logic behind such RAM-per-node values? What symptoms usually
suggest you that you need more RAM?
Did you ever get C* node soft lockdowns/not-respondings due to node being
loaded up to 100% of either ram/cpu/IO? If yes - under which conditions?
@Jeff Jirsa
>>Should not need regular reboots for Cassandra’s sake, but you should be
regularly patching your servers
Do you mean system updates/OS tuning and corresponding reloads, or some OS
cache cleanups?
Best regards,
Vsevolod.
вт, 17 июл. 2018 г., 17:18 Jeff Jirsa :
> Should not need reg
The default superuser is only created at startup if 3 conditions are met:
i) The default role manager is configured. In cassandra.yaml, you should
see "role_manager: CassandraRoleManager". This is also the default value,
so unless you're explicitly using a custom role manager it should be good.
ii
Very strange. CassandraRoleManager will create cassandra user if it doesn't
already exist (look in the setupDefaultRole method).
Perhaps this is an issue with consistency. After starting Cassandra, could you
try do a nodetool repair on system_auth keyspace on all nodes.
I'm starting to run out of
Thanks for your reply,
- I have not defined role_manager in the config
- I dropped the users table, it was present in the keyspace
- Cassandra then created a record in the roles table, yay !
But when I do clash -u cassandra -p cassandra
=> Invalid credentials supplied.
Authentication error on ho
Hi list,
Cassandra 3.0.6
I'd like to test the change of concurrent compactors to see if it helps
when the system is under stress.
Can someone point me to the right mbean?
I can not really find good docs about mbeans (or tools ...)
Any suggestion much appreciated, best
nodetool tablestats has an attribute for Percent repaired, can we target the
tables based the % given.
> On Jul 17, 2018, at 4:45 AM, Rahul Singh wrote:
>
> Have you considered looking into reaper project — could save you time in
> figuring out your own strategy.
> https://github.com/thelas
I would recommend at least looking at reaper before trying to engineer another
way.
On Jul 17, 2018, 12:56 PM -0400, rajasekhar kommineni ,
wrote:
> nodetool tablestats has an attribute for Percent repaired, can we target the
> tables based the % given.
>
>
>
> > On Jul 17, 2018, at 4:45 AM, Rah
If you have a read-heavy cluster, even with LCS you can still optimize by
having more of the key / row cache in memory.
If you have a write heavy / read heavy , then you need more memory so that more
data is available in the memtable as its written.
Not having enough memory means not enough hea
Hello Riccardo,
I noticed I have been writing a novel to answer a simple couple of
questions again ¯\_(ツ)_/¯. So here is a short answer in the case that's
what you were looking for :). Also, there is a warning that it might be
counter-productive and stress the cluster even more to increase the
com
Cassandra allows to use custom authenticators so I would create a
CustomPasswordAuthenticator. This would be a copy of the existing
PasswordAuthenticator. I would add several debugging info like: provided
username and password, the output of the checkpw function, what cql statement
is executed
Can I ask you an additional question here?
How much free space should I have if most tables use
TimeWindowCompactionStrategy?
> On 13 Jul 2018, at 10:09 PM, Vitaliy Semochkin wrote:
>
> Jeff, thank you very much for reply.
> Will try to use 4TB per instance.
>
> If I understand it correctly
Depends on whether or not you’re seeing overlaps (do you have read repairs
messing up the windows? Do you expire cleanly? Do you do non-ttl deletes?)
When I wrote it I used to run about 70-80% full, but we were willing to do some
manual cleanup when overlaps prevented dropping full sstables
--
I actually never set Xmx > 32 GB, for any java application, unless it
necessarily need more. Just because of the fact: "once you exceed this 32 GiB
border JVM will stop using compressed object pointers, effectively reducing the
available memory. That means increasing your JVM heap above 32 GiB y
Unfortunately, I’m not a java dev so I’m not able to create an authenticator…
I don’t like to do that usually but I share with you a gist of the config, it
was generated by OpsCenter when it was free, I just updated it for Cassandra >=
3… Maybe you will see something :
https://gist.github.com/b
If this is the file that you are currently using...he first things that
I see is that you do not have any authenticator and role_manager:
https://github.com/apache/cassandra/blob/1d506f9d09c880ff2b2693e3e27fa5
8c02ecf398/conf/cassandra.yaml#L103
https://github.com/apache/cassandra/blob/1d506f9d09
42 matches
Mail list logo
