Loopback Addresses

Every system has entire range of loopback address support, that is from 127.0.0.1 to 127.255.255.255

We can conform this by trying

ping -c 1 127.0.0.1

ping -c 1 127.0.0.2

Here, -c is count where were are fetching only 1 value.

***

Default Ports for Datastax/Cassandra

22         ssh

4040 Spark application website port

5598, 5599  Public/internode for DSE FS

7080 Spark Master console port

7081 Spark Worker Website port

8182 Gremlin server port for DSE Graph

8983 Solr port

8090 Spark jobserver REST API port

9042 Native Transport port

9091 Datastax Studio server port

9077  AlwaysOn SQL webui port

9142 DSE client port when SSL is Enabled

9999 Spark Jobserver JMX Port

18080 Spark application history server port

8888 Opscenter Port

7000 DSE internode cluster communication

7001 DSE SSL internode cluster communicaton

7199 DSE JMX metrics port

7077 Spark master internode communication

8609 Internode-messaging service port

8443 Opscenter SSL port

61619 Opscenter Server/stomp port

61620 Datastax agents for stomp server communications

61621 Datastax agents for http/https service

3128  LCM targets

***

DSEFS Errors

Cause: DSEFS was enabled and used the default keyspace across different DC's

Steps in resolving the Issue:

1) From dse fs unmount the existing locations.

2) Disable dsefs in dse.yaml and restart all nodes in the DC one at a time.

3) Drop corresponding dsefs keyspace.

4) Change dse.yaml configurations with new parameters for keyspace, directory paths etc

5) Restart service on all nodes one at a time.

6) Change the Replication strategy and Replication factor of the new dsefs keyspace.

7) Do nodetool repair on keyspace.

***

Creating ext4 or xfs file system in EC2 for /dev/xvd*

Identify the disks by verifying with,

lsblk

ls /dev/xvd*

Use below command to know about device, and its file system. If the output shows simply data then there is no file system on the device.

file -s /dev/xvdd

/dev/xvdd: data

or If the device has a file system, it will show the file system type, as example below.

file -s /dev/xvdd

/dev/xvdd: Linux rev 1.0 ext4 filesystem data, UUID=5b906ad5-f67a-474e-a7d4-7787586c72a3 (needs journal recovery) (extents) (64bit) (large files) (huge files)

--------------------------------------------------

If there is no file system then you can create it by

mkfs.ext4 /dev/xvdd

mkdir -p /cassandra/data1

mount /dev/xvdd /cassandra/data1

or 

mkfs -t xfs /dev/xvdf

If there is no xfs installed do it by "yum install xfsprogs"

mkdir -p /cassandra/commitlog

mount /dev/xvdf /cassandra/commitlog

----------------------------------------------------

Make the changes persistent by adding the entires in /etc/fstab

as example:vi /etc/fstab

/dev/xvdd /cassandra/data1 ext4 defaults 0 0

/dev/xvdf /cassandra/commitlog ext4 defaults 0 0

then on command prompt

do

mount -a (refresh the file system)

***

Compaction Threshold Settings in Cassandra

If there were multiple SSTables, read operations are less efficient as columns for associated key may be spread over multiple SSTables.

Adjusting Min and Max compaction thresholds helps in reducing frequent compactions but when it does, more SSTables are compacted.

nodetool getcompactionthreshold

Will give you current min = 4 and max = 32 for all tables which is default.

Change min and max threshold to 5 and 30 as below

nodetool setcompactionthreshold 5 30

Then alter the corresponding tables min_threshold and max_threshold accordingly as in this case to 5 and 30.

min_threshold, will compact minimum number of like sized SSTables and max_threshold controls the maximum number of SSTables that exist before minor compaction is forced. 

***

SAR: System Activity Reporter

SAR: System Activity Reporter(Available in systat package)

SAR is useful in understanding previous 10days statistics also, implies in go back in time.

sadc is the backend data collector for SAR.

SAR comes with two tools called sa1 and sa2

sa1: every 10 minutes(binary report)

sa2: summarize collected data each day(text file)

Both sa1 and sa2(sar) files will be loacted under /var/log/sa

Config file locations:

Ubuntu: /etc/default/sysstat - enable to true

CentOS: /etc/sysconfig/sysstat

SAR is basically a cron jobs

Ubuntu: /etc/cron.d/sysstat

Fedora systemd timer: /etc/systemd/system/sysstat.service.wants/sysstat-collect.timer (for sa1) & summary.timer (for sa2)

***