Google Git
Sign in
apache / cassandra / refs/heads/cassandra-2.0 / . / NEWS.txt
blob: 4b3fdf4dcd399e28e67d12b0442200d912842d8e [file] [log] [blame]
GENERAL UPGRADING ADVICE FOR ANY VERSION
========================================
Snapshotting is fast (especially if you have JNA installed) and takes
effectively zero disk space until you start compacting the live data
files again. Thus, best practice is to ALWAYS snapshot before any
upgrade, just in case you need to roll back to the previous version.
(Cassandra version X + 1 will always be able to read data files created
by version X, but the inverse is not necessarily the case.)
When upgrading major versions of Cassandra, you will be unable to
restore snapshots created with the previous major version using the
'sstableloader' tool. You can upgrade the file format of your snapshots
using the provided 'sstableupgrade' tool.
2.0.17
======
Upgrading
---------
- Nothing specific to this release, but refer to previous entries if you
are upgrading from a previous version.
2.0.16
======
Upgrading
---------
- Nothing specific to this release, but refer to previous entries if you
are upgrading from a previous version.
2.0.15
======
Upgrading
---------
- Nothing specific to this release, but refer to previous entries if you
are upgrading from a previous version.
2.0.14
======
The default JMX config now listens to localhost only. You must enable
the other JMX flags in cassandra-env.sh manually.
2.0.13
======
Upgrading
---------
- Prepending a list to a list collection was erroneously resulting in
the prepended list being reversed upon insertion. If you were depending
on this buggy behavior, note that it has been corrected.
2.0.12
======
Upgrading
---------
- Nothing specific to this release, but refer to previous entries if you
are upgrading from a previous version.
2.0.11
======
Upgrading
---------
- Nothing specific to this release, but refer to previous entries if you
are upgrading from a previous version.
New features
------------
- DateTieredCompactionStrategy added, optimized for time series data and groups
data that is written closely in time (CASSANDRA-6602 for details). Consider
this experimental for now.
2.0.10
======
New features
------------
- CqlPaginRecordReader and CqlPagingInputFormat have both been removed.
Use CqlInputFormat instead.
- If you are using Leveled Compaction, you can now disable doing size-tiered
compaction in L0 by starting Cassandra with -Dcassandra.disable_stcs_in_l0
(see CASSANDRA-6621 for details).
- Shuffle and taketoken have been removed. For clusters that choose to
upgrade to vnodes, creating a new datacenter with vnodes and migrating is
recommended. See http://goo.gl/Sna2S1 for further information.
2.0.9
=====
Upgrading
---------
- Default values for read_repair_chance and local_read_repair_chance have been
swapped. Namely, default read_repair_chance is now set to 0.0, and default
local_read_repair_chance to 0.1.
- Queries selecting only CQL static columns were (mistakenly) not returning one
result per row in the partition. This has been fixed and a SELECT DISTINCT
can be used when only the static column of a partition needs to be fetch
without fetching the whole partition. But if you use static columns, please
make sure this won't affect you (see CASSANDRA-7305 for details).
2.0.8
=====
New features
------------
- New snitches have been used for users of Google Compute Engine and of
Cloudstack.
Upgrading
---------
- Nothing specific to this release, but please see 2.0.7 if you are upgrading
from a previous version.
2.0.7
=====
Upgrading
---------
- Nothing specific to this release, but please see 2.0.6 if you are upgrading
from a previous version.
2.0.6
=====
New features
------------
- CQL now support static columns, allows to batch multiple conditional updates
and has a new syntax for slicing over multiple clustering columns
(http://goo.gl/B6qz4j).
- Repair can be restricted to a set of nodes using the -hosts option in nodetool.
- A new 'nodetool taketoken' command relocate tokens with vnodes.
- Hinted handoff can be enabled only for some data-centers (see
hinted_handoff_enabled in cassandra.yaml)
Upgrading
---------
- Nothing specific to this release, but please see 2.0.5 if you are upgrading
from a previous version.
2.0.5
=====
New features
------------
- Batchlog replay can be, and is throttled by default now.
See batchlog_replay_throttle_in_kb setting in cassandra.yaml.
- Scrub can now optionally skip corrupt counter partitions. Please note
that this will lead to the loss of all the counter updates in the skipped
partition. See the --skip-corrupted option.
Upgrading
---------
- If your cluster began on a version before 1.2, check that your secondary
index SSTables are on version 'ic' before upgrading. If not, run
'nodetool upgradesstables' if on 1.2.14 or later, or run 'nodetool
upgradesstables ks cf' with the keyspace and secondary index named
explicitly otherwise. If you don't do this and upgrade to 2.0.x and it
refuses to start because of 'hf' version files in the secondary index,
you will need to delete/move them out of the way and recreate the index
when 2.0.x starts.
2.0.3
=====
New features
------------
- It's now possible to configure the maximum allowed size of the native
protocol frames (native_transport_max_frame_size_in_mb in the yaml file).
Upgrading
---------
- NaN and Infinity are new valid floating point constants in CQL3 and are now reserved
keywords. In the unlikely case you were using one of them as an identifier (for a
column, a keyspace or a table), you will now have to double-quote them (see
http://cassandra.apache.org/doc/cql3/CQL.html#identifiers for "quoted identifiers").
- The IEndpointStateChangeSubscriber has a new method, beforeChange, that
any custom implemenations using the class will need to implement.
2.0.2
=====
New features
------------
- Speculative retry defaults to 99th percentile
(See blog post at http://www.datastax.com/dev/blog/rapid-read-protection-in-cassandra-2-0-2)
- Configurable metrics reporting
(see conf/metrics-reporter-config-sample.yaml)
- Compaction history and stats are now saved to system keyspace
(system.compaction_history table). You can access historiy via
new 'nodetool compactionhistory' command or CQL.
Upgrading
---------
- Nodetool defaults to Sequential mode for repair operations
2.0.1
=====
Upgrading
---------
- The default memtable allocation has changed from 1/3 of heap to 1/4
of heap. Also, default (single-partition) read and write timeouts
have been reduced from 10s to 5s and 2s, respectively.
2.0.0
=====
Upgrading
---------
- Java 7 is now *required*!
- Upgrading is ONLY supported from Cassandra 1.2.9 or later. This
goes for sstable compatibility as well as network. When
upgrading from an earlier release, upgrade to 1.2.9 first and
run upgradesstables before proceeding to 2.0.
- CAS and new features in CQL such as DROP COLUMN assume that cell
timestamps are microseconds-since-epoch. Do not use these
features if you are using client-specified timestamps with some
other source.
- Replication and strategy options do not accept unknown options anymore.
This was already the case for CQL3 in 1.2 but this is now the case for
thrift too.
- auto_bootstrap of a single-token node with no initial_token will
now pick a random token instead of bisecting an existing token
range. We recommend upgrading to vnodes; failing that, we
recommend specifying initial_token.
- reduce_cache_sizes_at, reduce_cache_capacity_to, and
flush_largest_memtables_at options have been removed from cassandra.yaml.
- CacheServiceMBean.reduceCacheSizes() has been removed.
Use CacheServiceMBean.set{Key,Row}CacheCapacityInMB() instead.
- authority option in cassandra.yaml has been deprecated since 1.2.0,
but it has been completely removed in 2.0. Please use 'authorizer' option.
- ASSUME command has been removed from cqlsh. Use CQL3 blobAsType() and
typeAsBlob() conversion functions instead.
See https://cassandra.apache.org/doc/cql3/CQL.html#blobFun for details.
- Inputting blobs as string constants is now fully deprecated in
favor of blob constants. Make sure to update your applications to use
the new syntax while you are still on 1.2 (which supports both string
and blob constants for blob input) before upgrading to 2.0.
- index_interval is now moved to ColumnFamily property. You can change value
with ALTER TABLE ... WITH statement and SSTables written after that will
have new value. When upgrading, Cassandra will pick up the value defined in
cassanda.yaml as the default for existing ColumnFamilies, until you explicitly
set the value for those.
- The deprecated native_transport_min_threads option has been removed in
Cassandra.yaml.
Operations
----------
- VNodes are enabled by default in cassandra.yaml. initial_token
for non-vnode deployments has been removed from the example
yaml, but is still respected if specified.
- Major compactions, cleanup, scrub, and upgradesstables will interrupt
any in-progress compactions (but not repair validations) when invoked.
- Disabling autocompactions by setting min/max compaction threshold to 0
has been deprecated, instead, use the nodetool commands 'disableautocompaction'
and 'enableautocompaction' or set the compaction strategy option enabled = false
- ALTER TABLE DROP has been reenabled for CQL3 tables and has new semantics now.
See https://cassandra.apache.org/doc/cql3/CQL.html#alterTableStmt and
https://issues.apache.org/jira/browse/CASSANDRA-3919 for details.
- CAS uses gc_grace_seconds to determine how long to keep unused paxos
state around for, or a minimum of three hours.
- A new hints created metric is tracked per target, replacing countPendingHints
- After performance testing for CASSANDRA-5727, the default LCS filesize
has been changed from 5MB to 160MB.
- cqlsh DESCRIBE SCHEMA no longer outputs the schema of system_* keyspaces;
use DESCRIBE FULL SCHEMA if you need the schema of system_* keyspaces.
- CQL2 has been deprecated, and will be removed entirely in 2.2. See
CASSANDRA-5918 for details.
- Commit log archiver now assumes the client time stamp to be in microsecond
precision, during restore. Please refer to commitlog_archiving.properties.
Features
--------
- Lightweight transactions
(http://www.datastax.com/dev/blog/lightweight-transactions-in-cassandra-2-0)
- Alias support has been added to CQL3 SELECT statement. Refer to
CQL3 documentation (http://cassandra.apache.org/doc/cql3/CQL.html) for details.
- JEMalloc support (see memory_allocator in cassandra.yaml)
- Experimental triggers support. See examples/ for how to use. "Experimental"
means "tied closely to internal data structures; we plan to decouple this in
the future, which will probably break triggers written against this initial
API."
- Numerous improvements to CQL3 and a new version of the native protocol. See
http://www.datastax.com/dev/blog/cql-in-cassandra-2-0 for details.
1.2.11
======
Features
--------
- Added a new consistency level, LOCAL_ONE, that forces all CL.ONE operations to
execute only in the local datacenter.
- New replace_address to supplant the (now removed) replace_token and
replace_node workflows to replace a dead node in place. Works like the
old options, but takes the IP address of the node to be replaced.
1.2.9
=====
Features
--------
- A history of executed nodetool commands is now captured.
It can be found in ~/.cassandra/nodetool.history. Other tools output files
(cli and cqlsh history, .cqlshrc) are now centralized in ~/.cassandra, as well.
- A new sstablesplit utility allows to split large sstables offline.
1.2.8
=====
Upgrading
---------
- Nothing specific to this release, but please see 1.2.7 if you are upgrading
from a previous version.
1.2.7
=====
Upgrading
---------
- If you have decommissioned a node in the past 72 hours, it is imperative
that you not upgrade until such time has passed, or do a full cluster
restart (not rolling) before beginning the upgrade. This only applies to
decommission, not removetoken.
1.2.6
=====
Upgrading
---------
- hinted_handoff_throttle_in_kb is now reduced by a factor
proportional to the number of nodes in the cluster (see
https://issues.apache.org/jira/browse/CASSANDRA-5272).
- CQL3 syntax for CREATE CUSTOM INDEX has been updated. See CQL3
documentation for details.
1.2.5
=====
Features
--------
- Custom secondary index support has been added to CQL3. Refer to
CQL3 documentation (http://cassandra.apache.org/doc/cql3/CQL.html)
for details and examples.
Upgrading
---------
- The native CQL transport is enabled by default on part 9042.
1.2.4
=====
Upgrading
---------
- 'nodetool upgradesstables' now only upgrades/rewrites sstables that are
not on the current version (which is usually what you want). Use the new
-a flag to recover the old behavior of rewriting all sstables.
Features
--------
- superuser setup delay (10 seconds) can now be overridden using
'cassandra.superuser_setup_delay_ms' property.
1.2.3
=====
Upgrading
---------
- CQL3 used to be case-insensitive for property map key in ALTER and CREATE
statements. In other words:
CREATE KEYSPACE test WITH replication = { 'CLASS' : 'SimpleStrategy',
'REPLICATION_FACTOR' : '1' }
was allowed. However, this was not consistent with the fact that string
literal are case sensitive in every other places and more importantly this
break NetworkTopologyStrategy for which DC names are case sensitive. Those
property map key are now case sensitive. So the statement above should be
changed to:
CREATE KEYSPACE test WITH replication = { 'class' : 'SimpleStrategy',
'replication_factor' : '1' }
1.2.2
=====
Upgrading
---------
- CQL3 type validation for constants has been fixed, which may require
fixing queries that were relying on the previous loose validation. Please
refer to the CQL3 documentation (http://cassandra.apache.org/doc/cql3/CQL.html)
and in particular the changelog section for more details. Please note in
particular that inputing blobs as strings constants is now deprecated (in
favor of blob constants) and its support will be removed in a future
version.
Features
--------
- Built-in CQL3-based implementations of IAuthenticator (PasswordAuthenticator)
and IAuthorizer (CassandraAuthorizer) have been added. PasswordAuthenticator
stores usernames and hashed passwords in system_auth.credentials table;
CassandraAuthorizer stores permissions in system_auth.permissions table.
- system_auth keyspace is now alterable via ALTER KEYSPACE queries.
The default is SimpleStrategy with replication_factor of 1, but it's
advised to raise RF to at least 3 or 5, since CL.QUORUM is used for all
auth-related queries. It's also possible to change the strategy to NTS.
- Permissions caching with time-based expiration policy has been added to reduce
performance impact of authorization. Permission validity can be configured
using 'permissions_validity_in_ms' setting in cassandra.yaml. The default
is 2000 (2 seconds).
- SimpleAuthenticator and SimpleAuthorizer examples have been removed. Please
look at CassandraAuthorizer/PasswordAuthenticator instead.
1.2.1
=====
Upgrading
---------
- In CQL3, date string are no longer accepted as timeuuid value since a
date string is not a correct representation of a timeuuid. Instead, new
methods (minTimeuuid, maxTimeuuid, now, dateOf, unixTimestampOf) have been
introduced to make working on timeuuid from date string easy. cqlsh also
does not display timeuuid as date string (since this is a lossy
representation), but the new dateOf method can be used instead. Please
refer to the reference documentation (http://cassandra.apache.org/doc/cql3/CQL.html)
for more detail.
- For client implementors: CQL3 client using the thrift interface should
use the new execute_cql3_query, prepare_cql3_query and execute_prepared_cql3_query
since 1.2.0. However, Cassandra 1.2.0 was not complaining if CQL3 was set
through set_cql_version but the now CQL2 only methods were used. This is
now the case.
- Queries that uses unrecognized or bad compaction or replication strategy
options are now refused (instead of simply logging a warning).
1.2
===
Upgrading
---------
- IAuthenticator interface has been updated to support dynamic
user creation, modification and removal. Users, even when stored
externally, now have to be explicitly created using
CREATE USER query first. AllowAllAuthenticator and SimpleAuthenticator
have been updated for the new interface, but you'll have to update
your old IAuthenticator implementations for 1.2. To ease this process,
a new abstract LegacyAuthenticator class has been added - subclass it
in your old IAuthenticator implementaion and everything should just work
(this only affects users who implemented custom authenticators).
- IAuthority interface has been deprecated in favor of IAuthorizer.
AllowAllAuthority and SimpleAuthority have been renamed to
AllowAllAuthorizer and SimpleAuthorizer, respectively. In order to
simplify the upgrade to the new interface, a new abstract
LegacyAuthorizer has been added - you should subclass it in your
old IAuthority implementation and everything should just work
(this only affects users who implemented custom authorities).
'authority' setting in cassandra.yaml has been renamed to 'authorizer',
'authority' is no longer recognized. This affects all upgrading users.
- 1.2 is NOT network-compatible with versions older than 1.0. That
means if you want to do a rolling, zero-downtime upgrade, you'll need
to upgrade first to 1.0.x or 1.1.x, and then to 1.2. 1.2 retains
the ability to read data files from Cassandra versions at least
back to 0.6, so a non-rolling upgrade remains possible with just
one step.
- The default partitioner for new clusters is Murmur3Partitioner,
which is about 10% faster for index-intensive workloads. Partitioners
cannot be changed once data is in the cluster, however, so if you are
switching to the 1.2 cassandra.yaml, you should change this to
RandomPartitioner or whatever your old partitioner was.
- If you using counters and upgrading from a version prior to
1.1.6, you should drain existing Cassandra nodes prior to the
upgrade to prevent overcount during commitlog replay (see
CASSANDRA-4782). For non-counter uses, drain is not required
but is a good practice to minimize restart time.
- Tables using LeveledCompactionStrategy will default to not
creating a row-level bloom filter. The default in older versions
of Cassandra differs; you should manually set the false positive
rate to 1.0 (to disable) or 0.01 (to enable, if you make many
requests for rows that do not exist).
- The hints schema was changed from 1.1 to 1.2. Cassandra automatically
snapshots and then truncates the hints column family as part of
starting up 1.2 for the first time. Additionally, upgraded nodes
will not store new hints destined for older (pre-1.2) nodes. It is
therefore recommended that you perform a cluster upgrade when all
nodes are up. Because hints will be lost, a cluster-wide repair (with
-pr) is recommended after upgrade of all nodes.
- The `nodetool removetoken` command (and corresponding JMX operation)
have been renamed to `nodetool removenode`. This function is
incompatible with the earlier `nodetool removetoken`, and attempts to
remove nodes in this way with a mixed 1.1 (or lower) / 1.2 cluster,
is not supported.
- The somewhat ill-conceived CollatingOrderPreservingPartitioner
has been removed. Use Murmur3Partitioner (recommended) or
ByteOrderedPartitioner instead.
- Global option hinted_handoff_throttle_delay_in_ms has been removed.
hinted_handoff_throttle_in_kb has been added instead.
- The default bloom filter fp chance has been increased to 1%.
This will save about 30% of the memory used by the old default.
Existing columnfamilies will retain their old setting.
- The default partitioner (for new clusters; the partitioner cannot be
changed in existing clusters) was changed from RandomPartitioner to
Murmur3Partitioner which provides faster hashing as well as improved
performance with secondary indexes.
- The default version of CQL (and cqlsh) is now CQL3. CQL2 is still
available but you will have to use the thrift set_cql_version method
(that is already supported in 1.1) to use CQL2. For cqlsh, you will need
to use 'cqlsh -2'.
- CQL3 is now considered final in this release. Compared to the beta
version that is part of 1.1, this final version has a few additions
(collections), but also some (incompatible) changes in the syntax for the
options of the create/alter keyspace/table statements. Typically, the
syntax to create a keyspace is now:
CREATE KEYSPACE ks WITH replication = { 'class' : 'SimpleStrategy',
'replication_factor' : 2 };
Also, the consistency level cannot be set in the language anymore, but is
at the protocol level.
Please refer to the CQL3 documentation (http://cassandra.apache.org/doc/cql3/CQL.html)
for details.
- In CQL3, the DROP behavior from ALTER TABLE has currently been removed
(because it was not correctly implemented). We hope to add it back soon
(Cassandra 1.2.1 or 1.2.2)
Features
--------
- Cassandra can now handle concurrent CREATE TABLE schema changes
as well as other updates
- rpc_timeout has been split up to allow finer-grained control
on timeouts for different operation types
- num_tokens can now be specified in cassandra.yaml. This defines the
number of tokens assigned to the host on the ring (default: 1).
Also specifying initial_token will override any num_tokens setting.
- disk_failure_policy allows blacklisting failed disks in JBOD
configuration instead of erroring out indefinitely
- event tracing can be configured per-connection ("trace_next_query")
or globally/probabilistically ("nodetool settraceprobability")
- Atomic batches are now supported server side, where Cassandra will
guarantee that (at the price of pre-writing the batch to another node
first), all mutations in the batch will be applied, even if the
coordinator fails mid-batch.
- new IAuthorizer interface has replaced the old IAuthority. IAuthorizer
allows dynamic permission management via new CQL3 statements:
GRANT, REVOKE, LIST PERMISSIONS. A native implementation storing
the permissions in Cassandra is being worked on and we expect to
include it in 1.2.1 or 1.2.2.
- IAuthenticator interface has been updated to support dynamic user
creation, modification and removal via new CQL3 statements:
CREATE USER, ALTER USER, DROP USER, LIST USERS. A native implementation
that stores users in Cassandra itself is being worked on and is expected to
become part of 1.2.1 or 1.2.2.
1.1.5
=====
Upgrading
---------
- Nothing specific to this release, but please see 1.1 if you are upgrading
from a previous version.
1.1.4
=====
Upgrading
---------
- Nothing specific to this release, but please see 1.1 if you are upgrading
from a previous version.
1.1.3
=====
Upgrading
---------
- Running "nodetool upgradesstables" after upgrading is recommended
if you use Counter columnfamilies.
Features
--------
- the cqlsh COPY command can now export to CSV flat files
- added a new tools/bin/token-generator to facilitate generating evenly distributed tokens
1.1.2
=====
Upgrading
---------
- If you have column families using the LeveledCompactionStrategy, you should run scrub on those column families.
Features
--------
- cqlsh has a new COPY command to load data from CSV flat files
1.1.1
=====
Upgrading
---------
- Nothing specific to this release, but please see 1.1 if you are upgrading
from a previous version.
Features
--------
- Continuous commitlog archiving and point-in-time recovery.
See conf/commitlog_archiving.properties
- Incremental repair by token range, exposed over JMX
1.1
===
Upgrading
---------
- Compression is enabled by default on newly created ColumnFamilies
(and unchanged for ColumnFamilies created prior to upgrading).
- If you are running a multi datacenter setup, you should upgrade to
the latest 1.0.x (or 0.8.x) release before upgrading. Versions
0.8.8 and 1.0.3-1.0.5 generate cross-dc forwarding that is incompatible
with 1.1.
- EACH_QUORUM ConsistencyLevel is only supported for writes and will now
throw an InvalidRequestException when used for reads. (Previous
versions would silently perform a LOCAL_QUORUM read instead.)
- ANY ConsistencyLevel is only supported for writes and will now
throw an InvalidRequestException when used for reads. (Previous
versions would silently perform a ONE read for range queries;
single-row and multiget reads already rejected ANY.)
- The largest mutation batch accepted by the commitlog is now 128MB.
(In practice, batches larger than ~10MB always caused poor
performance due to load volatility and GC promotion failures.)
Larger batches will continue to be accepted but will not be
durable. Consider setting durable_writes=false if you really
want to use such large batches.
- Make sure that global settings: key_cache_{size_in_mb, save_period}
and row_cache_{size_in_mb, save_period} in conf/cassandra.yaml are
used instead of per-ColumnFamily options.
- JMX methods no longer return custom Cassandra objects. Any such methods
will now return standard Maps, Lists, etc.
- Hadoop input and output details are now separated. If you were
previously using methods such as getRpcPort you now need to use
getInputRpcPort or getOutputRpcPort depending on the circumstance.
- CQL changes:
+ Prior to 1.1, you could use KEY as the primary key name in some
select statements, even if the PK was actually given a different
name. In 1.1+ you must use the defined PK name.
- The sliced_buffer_size_in_kb option has been removed from the
cassandra.yaml config file (this option was a no-op since 1.0).
Features
--------
- Concurrent schema updates are now supported, with any conflicts
automatically resolved. Please note that simultaneously running
‘CREATE COLUMN FAMILY’ operation on the different nodes wouldn’t
be safe until version 1.2 due to the nature of ColumnFamily
identifier generation, for more details see CASSANDRA-3794.
- The CQL language has undergone a major revision, CQL3, the
highlights of which are covered at [1]. CQL3 is not
backwards-compatibile with CQL2, so we've introduced a
set_cql_version Thrift method to specify which version you want.
(The default remains CQL2 at least until Cassandra 1.2.) cqlsh
adds a --cql3 flag to enable this.
[1] http://www.datastax.com/dev/blog/schema-in-cassandra-1-1
- Row-level isolation: multi-column updates to a single row have
always been *atomic* (either all will be applied, or none)
thanks to the CommitLog, but until 1.1 they were not *isolated*
-- a reader may see mixed old and new values while the update
happens.
- Finer-grained control over data directories, allowing a ColumnFamily to
be pinned to specfic volume, e.g. one backed by SSD.
- The bulk loader is not longer a fat client; it can be run from an
existing machine in a cluster.
- A new write survey mode has been added, similar to bootstrap (enabled via
-Dcassandra.write_survey=true), but the node will not automatically join
the cluster. This is useful for cases such as testing different
compaction strategies with live traffic without affecting the cluster.
- Key and row caches are now global, similar to the global memtable
threshold. Manual tuning of cache sizes per-columnfamily is no longer
required.
- Off-heap caches no longer require JNA, and will work out of the box
on Windows as well as Unix platforms.
- Streaming is now multithreaded.
- Compactions may now be aborted via JMX or nodetool.
- The stress tool is not new in 1.1, but it is newly included in
binary builds as well as the source tree
- Hadoop: a new BulkOutputFormat is included which will directly write
SSTables locally and then stream them into the cluster.
YOU SHOULD USE BulkOutputFormat BY DEFAULT. ColumnFamilyOutputFormat
is still around in case for some strange reason you want results
trickling out over Thrift, but BulkOutputFormat is significantly
more efficient.
- Hadoop: KeyRange.filter is now supported with ColumnFamilyInputFormat,
allowing index expressions to be evaluated server-side to reduce
the amount of data sent to Hadoop.
- Hadoop: ColumnFamilyRecordReader has a wide-row mode, enabled via
a boolean parameter to setInputColumnFamily, that pages through
data column-at-a-time instead of row-at-a-time.
- Pig: can use the wide-row Hadoop support, by setting PIG_WIDEROW_INPUT
to true. This will produce each row's columns in a bag.
1.0.8
=====
Upgrading
---------
- Nothing specific to 1.0.8
Other
-----
- Allow configuring socket timeout for streaming
1.0.7
=====
Upgrading
---------
- Nothing specific to 1.0.7, please report to instruction for 1.0.6
Other
-----
- Adds new setstreamthroughput to nodetool to configure streaming
throttling
- Adds JMX property to get/set rpc_timeout_in_ms at runtime
- Allow configuring (per-CF) bloom_filter_fp_chance
1.0.6
=====
Upgrading
---------
- This release fixes an issue related to the chunk_length_kb option for
compressed sstables. If you use compression on some column families, it
is recommended after the upgrade to check the value for this option on
these column families (the default value is 64). In case the option would
not be set correctly, you should update the column family definition,
setting the right value and then run scrub on the column family.
- Please report to instruction for 1.0.5 if coming from an older version.
1.0.5
=====
Upgrading
---------
- 1.0.5 comes to fix two important regression of 1.0.4. So all information
concerning 1.0.4 are valid for this release, but please avoids upgrading
to 1.0.4.
1.0.4
=====
Upgrading
---------
- Nothing specific to 1.0.4 but please see the 1.0 upgrading section if
upgrading from a version prior to 1.0.0
Features
--------
- A new upgradesstables command has been added to nodetool. It is very
similar to scrub but without the ability to discard corrupted rows (and
as a consequence it does not snapshot automatically before). This new
command is to be prefered to scrub in all cases where sstables should be
rewritten to the current format for upgrade purposes.
JMX
---
- The path for the data, commit log and saved cache directories exposed
through JMX
- The in-memory bloom filter sizes are now exposed through JMX
1.0.3
=====
Upgrading
---------
- Nothing specific to 1.0.3 but please see the 1.0 upgrading section if
upgrading from a version prior to 1.0.0
Features
--------
- For non compressed sstables (compressed sstable already include more
fine grained checsums), a sha1 for the full sstable is now automatically
created (in a fix with suffix -Digest.sha1). It can be used to check the
sstable integrity with sha1sum.
1.0.2
=====
Upgrading
---------
- Nothing specific to 1.0.2 but please see the 1.0 upgrading section if
upgrading from a version prior to 1.0.0
Features
--------
- Cassandra CLI queries now have timing information
1.0.1
=====
Upgrading
---------
- If upgrading from a version prior to 1.0.0, please see the 1.0 Upgrading
section
- For running on Windows as a Service, procrun is no longer discributed
with Cassandra, see README.txt for more information on how to download
it if necessary.
- The name given to snapshots directories have been improved for human
readability. If you had scripts relying on it, you may need to update
them.
1.0
===
Upgrading
---------
- Upgrading from version 0.7.1+ or 0.8.2+ can be done with a rolling
restart, one node at a time. (0.8.0 or 0.8.1 are NOT network-compatible
with 1.0: upgrade to the most recent 0.8 release first.)
You do not need to bring down the whole cluster at once.
- After upgrading, run nodetool scrub against each node before running
repair, moving nodes, or adding new ones.
- CQL inserts/updates now generate microsecond resolution timestamps
by default, instead of millisecond. THIS MEANS A ROLLING UPGRADE COULD
MIX milliseconds and microseconds, with clients talking to servers
generating milliseconds unable to overwrite the larger microsecond
timestamps. If you are using CQL and this is important for your
application, you can either perform a non-rolling upgrade to 1.0, or
update your application first to use explicit timestamps with the "USING
timestamp=X" syntax.
- The BinaryMemtable bulk-load interface has been removed (use the
sstableloader tool instead).
- The compaction_thread_priority setting has been removed from
cassandra.yaml (use compaction_throughput_mb_per_sec to throttle
compaction instead).
- CQL types bytea and date were renamed to blob and timestamp, respectively,
to conform with SQL norms. CQL type int is now a 4-byte int, not 8
(which is still available as bigint).
- Cassandra 1.0 uses arena allocation to reduce old generation
fragmentation. This means there is a minimum overhead of 1MB
per ColumnFamily plus 1MB per index.
- The SimpleAuthenticator and SimpleAuthority classes have been moved to
the example directory (and are thus not available from the binary
distribution). They never provided actual security and in their current
state are only meant as examples.
Features
--------
- SSTable compression is supported through the 'compression_options'
parameter when creating/updating a column family. For instance, you can
create a column family Cf using compression (through the Snappy library)
in the CLI with:
create column family Cf with compression_options={sstable_compression: SnappyCompressor}
SSTable compression is not activated by default but can be activated or
deactivated at any time.
- Compressed SSTable blocks are checksummed to protect against bitrot
- New LevelDB-inspired compaction algorithm can be enabled by setting the
Columnfamily compaction_strategy=LeveledCompactionStrategy option.
Leveled compaction means you only need to keep a few MB of space free for
compaction instead of (in the worst case) 50%.
- Ability to use multiple threads during a single compaction. See
multithreaded_compaction in cassandra.yaml for more details.
- Windows Service ("cassandra.bat install" to enable)
- A dead node may be replaced in a single step by starting a new node
with -Dcassandra.replace_token=<token>. More details can be found at
http://wiki.apache.org/cassandra/Operations#Replacing_a_Dead_Node
- It is now possible to repair only the first range returned by the
partitioner for a node with `nodetool repair -pr`. It makes it
easier/possible to repair a full cluster without any work duplication by
running this command on every node of the cluster.
New data types
--------------
- decimal
Other
-----
- Hinted Handoff has two major improvements:
- Hint replay is much more efficient thanks to a change in the data model
- Hints are created for all replicas that do not ack a write. (Formerly,
only replicas known to be down when the write started were hinted.)
This means that running with read repair completely off is much more
viable than before, and the default read_repair_chance is reduced from 1.0
("always repair") to 0.1 ("repair 10% of the time").
- The old per-ColumnFamily memtable thresholds
(memtable_throughput_in_mb, memtable_operations_in_millions,
memtable_flush_after_mins) are ignored, in favor of the global
memtable_total_space_in_mb and commitlog_total_space_in_mb settings.
This does not affect client compatibility -- the old options are
still allowed, but have no effect. These options may be removed
entirely in a future release.
- Backlogged compactions will begin five minutes after startup. The 0.8
behavior of never starting compaction until a flush happens is usually
not what is desired, but a short grace period is useful to allow caches
to warm up first.
- The deletion of compacted data files is not performed during Garbage
Collection anymore. This means compacted files will now be deleted
without delay.
0.8.5
=====
Features
--------
- SSTables copied to a data directory can be loaded by a live node through
nodetool refresh (may be handy to load snapshots).
- The configured compaction throughput is exposed through JMX.
Other
-----
- The sstableloader is now bundled with the debian package.
- Repair detects when a participating node is dead and fails instead of
hanging forever.
0.8.4
=====
Upgrading
---------
- Nothing specific to 0.8.4
Other
-----
- This release comes to fix a bug in counter that could lead to
(important) over-count.
- It also fixes a slight upgrade regression from 0.8.3. It is thus advised
to jump directly to 0.8.4 if upgrading from before 0.8.3.
0.8.3
=====
Upgrading
---------
- Token removal has been revamped. Removing tokens in a mixed cluster with
0.8.3 will not work, so the entire cluster will need to be running 0.8.3
first, except for the dead node.
Features
--------
- It is now possible to use thrift asynchronous and
half-synchronous/half-asynchronous servers (see cassandra.yaml for more
details).
- It is now possible to access counter columns through Hadoop.
Other
-----
- This release fix a regression of 0.8 that can make commit log segment to
be deleted even though not all data it contains has been flushed.
Upgrades from 0.8.* is very much encouraged.
0.8.2
=====
Upgrading
---------
- 0.8.0 and 0.8.1 shipped with a bug that was setting the
replicate_on_write option for counter column families to false (this
option has no effect on non-counter column family). This is an unsafe
default and 0.8.2 correct this, the default for replicate_on_write is
now true. It is advised to update your counter column family definitions
if replicate_on_write was uncorrectly set to false (before or after
upgrade).
0.8.1
=====
Upgrading
---------
- 0.8.1 is backwards compatible with 0.8, upgrade can be achieved by a
simple rolling restart.
- If upgrading for earlier version (0.7), please refer to the 0.8 section
for instructions.
Features
--------
- Numerous additions/improvements to CQL (support for counters, TTL, batch
inserts/deletes, index dropping, ...).
- Add two new AbstractTypes (comparator) to support compound keys
(CompositeType and DynamicCompositeType), as well as a ReverseType to
reverse the order of any existing comparator.
- New option to bypass the commit log on some keyspaces (for advanced
users).
Tools
-----
- Add new data bulk loading utility (sstableloader).
0.8
===
Upgrading
---------
- Upgrading from version 0.7.1 or later can be done with a rolling
restart, one node at a time. You do not need to bring down the
whole cluster at once.
- After upgrading, run nodetool scrub against each node before running
repair, moving nodes, or adding new ones.
- Running nodetool drain before shutting down the 0.7 node is
recommended but not required. (Skipping this will result in
replay of entire commitlog, so it will take longer to restart but
is otherwise harmless.)
- 0.8 is fully API-compatible with 0.7. You can continue
to use your 0.7 clients.
- Avro record classes used in map/reduce and Hadoop streaming code have
been removed. Map/reduce can be switched to Thrift by changing
org.apache.cassandra.avro in import statements to
org.apache.cassandra.thrift (no class names change). Streaming support
has been removed for the time being.
- The loadbalance command has been removed from nodetool. For similar
behavior, decommission then rebootstrap with empty initial_token.
- Thrift unframed mode has been removed.
- The addition of key_validation_class means the cli will assume keys
are bytes, instead of strings, in the absence of other information.
See http://wiki.apache.org/cassandra/FAQ#cli_keys for more details.
Features
--------
- added CQL client API and JDBC/DBAPI2-compliant drivers for Java and
Python, respectively (see: drivers/ subdirectory and doc/cql)
- added distributed Counters feature;
see http://wiki.apache.org/cassandra/Counters
- optional intranode encryption; see comments around 'encryption_options'
in cassandra.yaml
- compaction multithreading and rate-limiting; see
'concurrent_compactors' and 'compaction_throughput_mb_per_sec' in
cassandra.yaml
- cassandra will limit total memtable memory usage to 1/3 of the heap
by default. This can be ajusted or disabled with the
memtable_total_space_in_mb option. The old per-ColumnFamily
throughput, operations, and age settings are still respected but
will be removed in a future major release once we are satisfied that
memtable_total_space_in_mb works adequately.
Tools
-----
- stress and py_stress moved from contrib/ to tools/
- clustertool was removed (see
https://issues.apache.org/jira/browse/CASSANDRA-2607 for examples
of how to script nodetool across the cluster instead)
Other
-----
- In the past, sstable2json would write column names and values as
hex strings, and now creates human readable values based on the
comparator/validator. As a result, JSON dumps created with
older versions of sstable2json are no longer compatible with
json2sstable, and imports must be made with a configuration that
is identical to the export.
- manually-forced compactions ("nodetool compact") will do nothing
if only a single SSTable remains for a ColumnFamily. To force it
to compact that anyway (which will free up space if there are
a lot of expired tombstones), use the new forceUserDefinedCompaction
JMX method on CompactionManager.
- most of contrib/ (which was not part of the binary releases)
has been moved either to examples/ or tools/. We plan to move the
rest for 0.8.1.
JMX
---
- By default, JMX now listens on port 7199.
0.7.6
=====
Upgrading
---------
- Nothing specific to 0.7.6, but see 0.7.3 Upgrading if upgrading
from earlier than 0.7.1.
0.7.5
=====
Upgrading
---------
- Nothing specific to 0.7.5, but see 0.7.3 Upgrading if upgrading
from earlier than 0.7.1.
Changes
-------
- system_update_column_family no longer snapshots before applying
the schema change. (_update_keyspace never did. _drop_keyspace
and _drop_column_family continue to snapshot.)
- added memtable_flush_queue_size option to cassandra.yaml to
avoid blocking writes when multiple column families (or a colum
family with indexes) are flushed at the same time.
- allow overriding initial_token, storage_port and rpc_port using
system properties
0.7.4
=====
Upgrading
---------
- Nothing specific to 0.7.4, but see 0.7.3 Upgrading if upgrading
from earlier than 0.7.1.
Features
--------
- Output to Pig is now supported as well as input
0.7.3
=====
Upgrading
---------
- 0.7.1 and 0.7.2 shipped with a bug that caused incorrect row-level
bloom filters to be generated when compacting sstables generated
with earlier versions. This would manifest in IOExceptions during
column name-based queries. 0.7.3 provides "nodetool scrub" to
rebuild sstables with correct bloom filters, with no data lost.
(If your cluster was never on 0.7.0 or earlier, you don't have to
worry about this.) Note that nodetool scrub will snapshot your
data files before rebuilding, just in case.
0.7.1
=====
Upgrading
---------
- 0.7.1 is completely backwards compatible with 0.7.0. Just restart
each node with the new version, one at a time. (The cluster does
not all need to be upgraded simultaneously.)
Features
--------
- added flush_largest_memtables_at and reduce_cache_sizes_at options
to cassandra.yaml as an escape valve for memory pressure
- added option to specify -Dcassandra.join_ring=false on startup
to allow "warm spare" nodes or performing JMX maintenance before
joining the ring
Performance
-----------
- Disk writes and sequential scans avoid polluting page cache
(requires JNA to be enabled)
- Cassandra performs writes efficiently across datacenters by
sending a single copy of the mutation and having the recipient
forward that to other replicas in its datacenter.
- Improved network buffering
- Reduced lock contention on memtable flush
- Optimized supercolumn deserialization
- Zero-copy reads from mmapped sstable files
- Explicitly set higher JVM new generation size
- Reduced i/o contention during saving of caches
0.7.0
=====
Features
--------
- Secondary indexes (indexes on column values) are now supported
- Row size limit increased from 2GB to 2 billion columns. rows
are no longer read into memory during compaction.
- Keyspace and ColumnFamily definitions may be added and modified live
- Streaming data for repair or node movement no longer requires
anticompaction step first
- NetworkTopologyStrategy (formerly DatacenterShardStrategy) is ready for
use, enabling ConsistencyLevel.DCQUORUM and DCQUORUMSYNC. See comments
in `cassandra.yaml.`
- Optional per-Column time-to-live field allows expiring data without
have to issue explicit remove commands
- `truncate` thrift method allows clearing an entire ColumnFamily at once
- Hadoop OutputFormat and Streaming [non-jvm map/reduce via stdin/out]
support
- Up to 8x faster reads from row cache
- A new ByteOrderedPartitioner supports bytes keys with arbitrary content,
and orders keys by their byte value. This should be used in new
deployments instead of OrderPreservingPartitioner.
- Optional round-robin scheduling between keyspaces for multitenant
clusters
- Dynamic endpoint snitch mitigates the impact of impaired nodes
- New `IntegerType`, faster than LongType and allows integers of
both less and more bits than Long's 64
- A revamped authentication system that decouples authorization and
allows finer-grained control of resources.
Upgrading
---------
The Thrift API has changed in incompatible ways; see below, and refer
to http://wiki.apache.org/cassandra/ClientOptions for a list of
higher-level clients that have been updated to support the 0.7 API.
The Cassandra inter-node protocol is incompatible with 0.6.x
releases (and with 0.7 beta1), meaning you will have to bring your
cluster down prior to upgrading: you cannot mix 0.6 and 0.7 nodes.
The hints schema was changed from 0.6 to 0.7. Cassandra automatically
snapshots and then truncates the hints column family as part of
starting up 0.7 for the first time.
Keyspace and ColumnFamily definitions are stored in the system
keyspace, rather than the configuration file.
The process to upgrade is:
1) run "nodetool drain" on _each_ 0.6 node. When drain finishes (log
message "Node is drained" appears), stop the process.
2) Convert your storage-conf.xml to the new cassandra.yaml using
"bin/config-converter".
3) Rename any of your keyspace or column family names that do not adhere
to the '^\w+' regex convention.
4) Start up your cluster with the 0.7 version.
5) Initialize your Keyspace and ColumnFamily definitions using
"bin/schematool <host> <jmxport> import". _You only need to do
this to one node_.
Thrift API
----------
- The Cassandra server now defaults to framed mode, rather than
unframed. Unframed is obsolete and will be removed in the next
major release.
- The Cassandra Thrift interface file has been updated for Thrift 0.5.
If you are compiling your own client code from the interface, you
will need to upgrade the Thrift compiler to match.
- Row keys are now bytes: keys stored by versions prior to 0.7.0 will be
returned as UTF-8 encoded bytes. OrderPreservingPartitioner and
CollatingOrderPreservingPartitioner continue to expect that keys contain
UTF-8 encoded strings, but RandomPartitioner now works on any key data.
- keyspace parameters have been replaced with the per-connection
set_keyspace method.
- The return type for login() is now AccessLevel.
- The get_string_property() method has been removed.
- The get_string_list_property() method has been removed.
Configuraton
------------
- Configuration file renamed to cassandra.yaml and log4j.properties to
log4j-server.properties
- PropertyFileSnitch configuration file renamed to
cassandra-topology.properties
- The ThriftAddress and ThriftPort directives have been renamed to
RPCAddress and RPCPort respectively.
- EndPointSnitch was renamed to RackInferringSnitch. A new SimpleSnitch
has been added.
- RackUnawareStrategy and RackAwareStrategy have been renamed to
SimpleStrategy and OldNetworkTopologyStrategy, respectively.
- RowWarningThresholdInMB replaced with in_memory_compaction_limit_in_mb
- GCGraceSeconds is now per-ColumnFamily instead of global
- Keyspace and column family names that do not confirm to a '^\w+' regex
are considered illegal.
- Keyspace and column family definitions will need to be loaded via
"bin/schematool <host> <jmxport> import". _You only need to do this to
one node_.
- In addition to an authenticator, an authority must be configured as
well. Users of SimpleAuthenticator should use SimpleAuthority for this
value (the default is AllowAllAuthority, which corresponds with
AllowAllAuthenticator).
- The format of access.properties has changed, see the sample configuration
conf/access.properties for documentation on the new format.
JMX
---
- StreamingService moved from o.a.c.streaming to o.a.c.service
- GMFD renamed to GOSSIP_STAGE
- {Min,Mean,Max}RowCompactedSize renamed to {Min,Mean,Max}RowSize
since it no longer has to wait til compaction to be computed
Other
-----
- If extending AbstractType, make sure you follow the singleton pattern
followed by Cassandra core AbstractType classes: provide a public
static final variable called 'instance'.
0.6.6
=====
Upgrading
---------
- As part of the cache-saving feature, a third directory
(along with data and commitlog) has been added to the config
file. You will need to set and create this directory
when restarting your node into 0.6.6.
0.6.1
=====
Upgrading
---------
- We try to keep minor versions 100% compatible (data format,
commitlog format, network format) within the major series, but
we introduced a network-level incompatibility in 0.6.1.
Thus, if you are upgrading from 0.6.0 to any higher version
(0.6.1, 0.6.2, etc.) then you will need to restart your entire
cluster with the new version, instead of being able to do a
rolling restart.
0.6.0
=====
Features
--------
- row caching: configure with the RowsCached attribute in
ColumnFamily definition
- Hadoop map/reduce support: see contrib/word_count for an example
- experimental authentication support, described under
Authenticator in storage.conf
Configuraton
------------
- MemtableSizeInMB has been replaced by MemtableThroughputInMB which
triggers a memtable flush when the specified amount of data has
been written, including overwrites.
- MemtableObjectCountInMillions has been replaced by the
MemtableOperationsInMillions directive which causes a memtable flush
to occur after the specified number of operations.
- Like MemtableSizeInMB, BinaryMemtableSizeInMB has been replaced by
BinaryMemtableThroughputInMB.
- Replication factor is now per-keyspace, rather than global.
- KeysCachedFraction is deprecated in favor of KeysCached
- RowWarningThresholdInMB added, to warn before very large rows
get big enough to threaten node stability
Thrift API
----------
- removed deprecated get_key_range method
- added batch_mutate meethod
- deprecated multiget and batch_insert methods in favor of
multiget_slice and batch_mutate, respectively
- added ConsistencyLevel.ANY, for when you want write
availability even when it may not be readable immediately.
Unlike CL.ZERO, though, it will throw an exception if
it cannot be written *somewhere*.
JMX metrics
-----------
- read and write statistics are reported as lifetime totals,
instead of averages over the last minute. average-since-last
requested are also available for convenience.
- cache hit rate statistics are now available from JMX under
org.apache.cassandra.db.Caches
- compaction JMX metrics are moved to
org.apache.cassandra.db.CompactionManager. PendingTasks is now
a much better estimate of compactions remaining, and the
progress of the current compaction has been added.
- commitlog JMX metrics are moved to org.apache.cassandra.db.Commitlog
- progress of data streaming during bootstrap, loadbalance, or other
data migration, is available under
org.apache.cassandra.streaming.StreamingService.
See http://wiki.apache.org/cassandra/Streaming for details.
Installation/Upgrade
--------------------
- 0.6 network traffic is not compatible with earlier versions. You
will need to shut down all your nodes at once, upgrade, then restart.
0.5.0
=====
0. The commitlog format has changed (but sstable format has not).
When upgrading from 0.4, empty the commitlog either by running
bin/nodeprobe flush on each machine and waiting for the flush to finish,
or simply remove the commitlog directory if you only have test data.
(If more writes come in after the flush command, starting 0.5 will error
out; if that happens, just go back to 0.4 and flush again.)
The format changed twice: from 0.4 to beta1, and from beta2 to RC1.
.5 The gossip protocol has changed, meaning 0.5 nodes cannot coexist
in a cluster of 0.4 nodes or vice versa; you must upgrade your
whole cluster at the same time.
1. Bootstrap, move, load balancing, and active repair have been added.
See http://wiki.apache.org/cassandra/Operations. When upgrading
from 0.4, leave autobootstrap set to false for the first restart
of your old nodes.
2. Performance improvements across the board, especially on the write
path (over 100% improvement in stress.py throughput).
3. Configuration:
- Added "comment" field to ColumnFamily definition.
- Added MemtableFlushAfterMinutes, a global replacement for the
old per-CF FlushPeriodInMinutes setting
- Key cache settings
4. Thrift:
- Added get_range_slice, deprecating get_key_range
0.4.2
=====
1. Improve default garbage collector options significantly --
throughput will be 30% higher or more.
0.4.1
=====
1. SnapshotBeforeCompaction configuration option allows snapshotting
before each compaction, which allows rolling back to any version
of the data.
0.4.0
=====
1. On-disk data format has changed to allow billions of keys/rows per
node instead of only millions. The new format is incompatible with 0.3;
see 0.3 notes below for how to import data from a 0.3 install.
2. Cassandra now supports multiple keyspaces. Typically you will have
one keyspace per application, allowing applications to be able to
create and modify ColumnFamilies at will without worrying about
collisions with others in the same cluster.
3. Many Thrift API changes and documentation. See
http://wiki.apache.org/cassandra/API
4. Removed the web interface in favor of JMX and bin/nodeprobe, which
has significantly enhanced functionality.
5. Renamed configuration "<Table>" to "<Keyspace>".
6. Added commitlog fsync; see "<CommitLogSync>" in configuration.
0.3.0
=====
1. With enough and large enough keys in a ColumnFamily, Cassandra will
run out of memory trying to perform compactions (data file merges).
The size of what is stored in memory is (S + 16) * (N + M) where S
is the size of the key (usually 2 bytes per character), N is the
number of keys and M, is the map overhead (which can be guestimated
at around 32 bytes per key).
So, if you have 10-character keys and 1GB of headroom in your heap
space for compaction, you can expect to store about 17M keys
before running into problems.
See https://issues.apache.org/jira/browse/CASSANDRA-208
2. Because fixing #1 requires a data file format change, 0.4 will not
be binary-compatible with 0.3 data files. A client-side upgrade
can be done relatively easily with the following algorithm:
for key in old_client.get_key_range(everything):
columns = old_client.get_slice or get_slice_super(key, all columns)
new_client.batch_insert or batch_insert_super(key, columns)
The inner loop can be trivially parallelized for speed.
3. Commitlog does not fsync before reporting a write successful.
Using blocking writes mitigates this to some degree, since all
nodes that were part of the write quorum would have to fail
before sync for data to be lost.
See https://issues.apache.org/jira/browse/CASSANDRA-182
Additionally, row size (that is, all the data associated with a single
key in a given ColumnFamily) is limited by available memory, because
compaction deserializes each row before merging.
See https://issues.apache.org/jira/browse/CASSANDRA-16