Well it’s hard to say unless you’re more precise with what JMX emitted metric 
you’re graphing, but yes, there are read metrics that will increase if people 
read the data


-- 
Jeff Jirsa


> On Mar 11, 2018, at 7:38 PM, Eunsu Kim <eunsu.bil...@gmail.com> wrote:
> 
> No I didn’t
> 
> Do you mean that this read throughput value comes out because somebody 
> actually reads the data?
> 
> 
>> On 12 Mar 2018, at 11:32 AM, Jeff Jirsa <jji...@gmail.com> wrote:
>> 
>> I presume you’re asking because you don’t think you’re doing reads. Did you 
>> start doing counter writes?
>> 
>> 
>> -- 
>> Jeff Jirsa
>> 
>> 
>>> On Mar 11, 2018, at 7:30 PM, Eunsu Kim <eunsu.bil...@gmail.com> wrote:
>>> 
>>> We monitored the write/read throughput through the Cassandra cluster via 
>>> JMX. There was little read throughput before March 6. From March 6, the 
>>> write throughput increased, but the read throughput suddenly increased 
>>> sharply. Do anyone know why this happens?
>>> 
>>> <Screen Shot 2018-03-12 at 10.27.07 AM.png>
> 

Reply via email to