Hello,
For the metrics
"org.apache.cassandra.metrics:name=[DC]-Latency,type=Messaging" which is
the unit of the metrics?
-b -i -q 75thPercentile
999thPercentile DurationUnit Max Min
StdDev
-d -l -s 95thPercentile
99thPercentile FifteenMinuteRate Mean
OneMinuteRate
-h -n 50thPercentile 98thPercentile
Count FiveMinuteRate MeanRate RateUnit
On Tue, Oct 17, 2017 at 10:27 PM Ben Wood <bw...@mesosphere.io> wrote:
> (https://www.cloudping.co/ is a useful place to get inter-region latency
> in AWS)
>
> On Tue, Oct 17, 2017 at 9:20 PM, Chris Lohfink <clohfin...@gmail.com>
> wrote:
>
>> An alternative if using >3.8 you can use the
>> org.apache.cassandra.metrics:type=Messaging,name=[DC]-Latency mbean where
>> [DC] is the name of the DC and you can get the inter DC latency per node
>> (to that node). This does not account for NTP drift though, just how long
>> it takes messages (ie mutations) take to get to a node from other DCs.
>>
>> Chris
>>
>> On Tue, Oct 17, 2017 at 7:18 PM, Jon Haddad <j...@jonhaddad.com> wrote:
>>
>>> I recommend figuring out the latency between your datacenters.
>>> Cassandra isn’t going to be any more than that barring JVM pauses on the
>>> remote coordinator.
>>>
>>>
>>> On Oct 17, 2017, at 4:17 PM, Bill Walters <billwalter...@gmail.com>
>>> wrote:
>>>
>>> Hi Everyone,
>>>
>>> I need some suggestions on finding the time taken for Cassandra
>>> replication to happen from east to west region for write and read
>>> operations on a multi DC cluster.
>>> Currently, below is our cluster setup.
>>>
>>> *Cassandra version:* DSE 5.0.7
>>> *No of Data centers:* 2 (AWS East and AWS West regions)
>>> *No of Nodes:* 12 nodes (6 nodes in AWS East and 6 nodes in AWS West)
>>> *Replication Factor:* 3 in each data center.
>>> *Cluster size*: Around 40 GB on each node
>>>
>>> Sometime, next year we have an activity where our clients are going to
>>> be reading only from AWS West region. The data center in AWS east will be
>>> available but we do not want any reads to be done on this.(Our management
>>> wants to know the time it takes for Cassandra to replicate from one DC to
>>> the other)
>>>
>>> Here are some options I have thought of in finding the time taken for
>>> Cassandra replication to happen from AWS East DC to AWS West DC.
>>>
>>> 1. Setup a Java client to write/read a transaction with *"Local Quorum"
>>> *consistency level in* AWS East* region as Local data center, capture
>>> the time taken for this activity. Similarly use this client to perform
>>> read/write transaction with *"Local Quorum"* consistency level in *AWS
>>> West* region and capture the time. Then finally perform the same
>>> transaction with with *"Each Quorum" *consistency level and capture the
>>> time.
>>>
>>> *Inter DC latency* = *Time taken for Each Quorum transaction* *-* *(Time
>>> taken for Local Quorum transaction in AWS East as local dc)* *-** (Time
>>> taken for Local Quorum transaction in AWS West as local dc)*.
>>>
>>>
>>> 2. Utilize the
>>> https://github.com/gitaroktato/cassandra-replication-latency-tools open
>>> source project where a Python Cassandra clients writes in one Data Center
>>> and other client reads in other data center.
>>>
>>>
>>> Can you please suggest if my strategies above will help in finding the
>>> Inter DC latency or there are other ways I need to follow.
>>>
>>>
>>> Thank You,
>>> Bill Walters.
>>>
>>>
>>>
>>
>
>
> --
> Ben Wood
> Software Engineer - Data Agility
> Mesosphere
>
