Re: Performance in solr7.5

2020-07-08 Thread Sankar Panda
Hi All,

Any suggestion?
Thanks
Sankar Panda

On Mon, Jul 6, 2020, 16:58 Sankar Panda  wrote:

> Hi Eric,
> Thanks for your mail. I am seeing that 91`% time cpu is idle.
>
> if you see below some of the stats ,still i am not able to visualize where
> it is wrong
>
> FilenameTypeSizeUsed
>  Priority
> /mnt/resource/swapfile  file2097148 1119016 -2
>
>  totalusedfree  shared  buff/cache   available
> Mem:251  51  16   2 184
>   196
> Swap: 1   1   0
>
>  vmstat
> procs ---memory-- ---swap-- -io -system--
> --cpu-
>  r  b   swpd   free   buff  cache   si   sobibo   in   cs us sy id
> wa st
>  0  0 1119016 16969144 200036 19279326400   177   21700  2
>  2 92  4  0
>
> avg-cpu:  %user   %nice %system %iowait  %steal   %idle
>2.490.001.604.140.00   91.77
>
> Can you please help me where i need to see
>
> Thanks
> Sankar Panda
>
>
>
> On Mon, Jul 6, 2020 at 12:19 AM Erick Erickson 
> wrote:
>
>> Look at your I/O stats. My bet is that you’re swapping like crazy and
>> your CPU is relatively idle.
>>
>> 2T of index on two machines is probably simply too much data on too
>> little hardware.
>>
>> Consider stress testing your hardware gradually, see:
>>
>>
>> https://lucidworks.com/post/sizing-hardware-in-the-abstract-why-we-dont-have-a-definitive-answer/
>>
>> Best,
>> Erick
>>
>> > On Jul 5, 2020, at 9:38 AM, Sankar Panda 
>> wrote:
>> >
>> > Hi All,
>> >
>> > I am facing performance issue while searching. It took 10mins to get the
>> > results.i have 2 shard and each shard having 2 replicas.80M documents
>> and
>> > index size in each shard is  2T.
>> >
>> > Any suggestions?
>> >
>> > Thanks
>> > Sankar panda
>>
>>


Re: Performance in solr7.5

2020-07-06 Thread Sankar Panda
Hi Eric,
Thanks for your mail. I am seeing that 91`% time cpu is idle.

if you see below some of the stats ,still i am not able to visualize where
it is wrong

FilenameTypeSizeUsed
 Priority
/mnt/resource/swapfile  file2097148 1119016 -2

 totalusedfree  shared  buff/cache   available
Mem:251  51  16   2 184
196
Swap: 1   1   0

 vmstat
procs ---memory-- ---swap-- -io -system--
--cpu-
 r  b   swpd   free   buff  cache   si   sobibo   in   cs us sy id
wa st
 0  0 1119016 16969144 200036 19279326400   177   21700  2
 2 92  4  0

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
   2.490.001.604.140.00   91.77

Can you please help me where i need to see

Thanks
Sankar Panda



On Mon, Jul 6, 2020 at 12:19 AM Erick Erickson 
wrote:

> Look at your I/O stats. My bet is that you’re swapping like crazy and your
> CPU is relatively idle.
>
> 2T of index on two machines is probably simply too much data on too little
> hardware.
>
> Consider stress testing your hardware gradually, see:
>
>
> https://lucidworks.com/post/sizing-hardware-in-the-abstract-why-we-dont-have-a-definitive-answer/
>
> Best,
> Erick
>
> > On Jul 5, 2020, at 9:38 AM, Sankar Panda  wrote:
> >
> > Hi All,
> >
> > I am facing performance issue while searching. It took 10mins to get the
> > results.i have 2 shard and each shard having 2 replicas.80M documents and
> > index size in each shard is  2T.
> >
> > Any suggestions?
> >
> > Thanks
> > Sankar panda
>
>


Re: Performance in solr7.5

2020-07-05 Thread Erick Erickson
Look at your I/O stats. My bet is that you’re swapping like crazy and your CPU 
is relatively idle.

2T of index on two machines is probably simply too much data on too little 
hardware.

Consider stress testing your hardware gradually, see:

https://lucidworks.com/post/sizing-hardware-in-the-abstract-why-we-dont-have-a-definitive-answer/

Best,
Erick

> On Jul 5, 2020, at 9:38 AM, Sankar Panda  wrote:
> 
> Hi All,
> 
> I am facing performance issue while searching. It took 10mins to get the
> results.i have 2 shard and each shard having 2 replicas.80M documents and
> index size in each shard is  2T.
> 
> Any suggestions?
> 
> Thanks
> Sankar panda



Performance in solr7.5

2020-07-05 Thread Sankar Panda
Hi All,

I am facing performance issue while searching. It took 10mins to get the
results.i have 2 shard and each shard having 2 replicas.80M documents and
index size in each shard is  2T.

Any suggestions?

Thanks
Sankar panda