Getting started with accumulo

[Frank Palladino <pa...@gmail.com>]

Hello Bigtop,

We have a working cluster which is managed by bigtop and would like to get
accumulo up and running on it. We have pulled down the accumulo branch on
github and made the packages, where do we go from here.

Any help would be appreciated.

Thanks in advance,

Frank Palladino

Re: Getting started with accumulo

[Frank Palladino <pa...@gmail.com>]

Thanks Jay and Sean,

I'll try to get it working and keep you posted on the progress.

Best

Frank


On Mon, Aug 18, 2014 at 6:55 PM, Sean Mackrory <ma...@gmail.com> wrote:

> If you've already made the packages, you should have an 'accumulo' package
> that provides all the common bits, and 'accumulo-doc' package that is not
> relevant to just getting the system up and running, and the following
> packages:
>
> accumulo-master
> accumulo-tserver
> accumulo-gc
> accumulo-monitor
> accumulo-tracer
>
> This final list of packages corresponds to the different daemons you can
> run. Installing those packages on the machines will get those roles to
> start upon reboot (or manually with 'service accumulo-master start', etc.)
> You can install the packages by either copying the .deb / .rpm files and
> installing them with 'dpkg -i' / 'rpm -i' respectively (depending on what
> OS you're running on). As with any other Bigtop package, configuration will
> be in /etc/accumulo and /etc/default/accumulo, logs will be in
> /var/log/accumulo, etc..
>
>
> On Sun, Aug 17, 2014 at 12:36 PM, jay vyas <ja...@gmail.com>
> wrote:
>
>> Hi frank.
>>
>> I think sean mackrory is the main guy involved with the accumulo stuff.
>>
>> - I would guess that the key thing is to grab your accumulo distribution
>> and confirm that it works with the exact version of hadoop you have (i.e.
>> hadoop version).
>>
>> - Sean do you have any thoughts on how to deploy accumulo packages?  I
>> dont think we have any puppet recipes for them yet.
>>
>> Will be interesting to  try out the accumulo branch at some point,
>> possibly adding smoke tests to them as well.
>>
>> - In the meantime, you can also check the accumulo mailing list, maybe
>> some of the folks there will be interested in testing / contributing to the
>> accumulo packages in bigtop.
>>
>>
>>
>>
>> On Fri, Aug 15, 2014 at 11:45 AM, Frank Palladino <
>> palladino.frank@gmail.com> wrote:
>>
>>> Hello Bigtop,
>>>
>>> We have a working cluster which is managed by bigtop and would like to
>>> get accumulo up and running on it. We have pulled down the accumulo branch
>>> on github and made the packages, where do we go from here.
>>>
>>> Any help would be appreciated.
>>>
>>> Thanks in advance,
>>>
>>> Frank Palladino
>>>
>>
>>
>>
>> --
>> jay vyas
>>
>
>

Re: Getting started with accumulo

[Sean Mackrory <ma...@gmail.com>]

If you've already made the packages, you should have an 'accumulo' package
that provides all the common bits, and 'accumulo-doc' package that is not
relevant to just getting the system up and running, and the following
packages:

accumulo-master
accumulo-tserver
accumulo-gc
accumulo-monitor
accumulo-tracer

This final list of packages corresponds to the different daemons you can
run. Installing those packages on the machines will get those roles to
start upon reboot (or manually with 'service accumulo-master start', etc.)
You can install the packages by either copying the .deb / .rpm files and
installing them with 'dpkg -i' / 'rpm -i' respectively (depending on what
OS you're running on). As with any other Bigtop package, configuration will
be in /etc/accumulo and /etc/default/accumulo, logs will be in
/var/log/accumulo, etc..


On Sun, Aug 17, 2014 at 12:36 PM, jay vyas <ja...@gmail.com>
wrote:

> Hi frank.
>
> I think sean mackrory is the main guy involved with the accumulo stuff.
>
> - I would guess that the key thing is to grab your accumulo distribution
> and confirm that it works with the exact version of hadoop you have (i.e.
> hadoop version).
>
> - Sean do you have any thoughts on how to deploy accumulo packages?  I
> dont think we have any puppet recipes for them yet.
>
> Will be interesting to  try out the accumulo branch at some point,
> possibly adding smoke tests to them as well.
>
> - In the meantime, you can also check the accumulo mailing list, maybe
> some of the folks there will be interested in testing / contributing to the
> accumulo packages in bigtop.
>
>
>
>
> On Fri, Aug 15, 2014 at 11:45 AM, Frank Palladino <
> palladino.frank@gmail.com> wrote:
>
>> Hello Bigtop,
>>
>> We have a working cluster which is managed by bigtop and would like to
>> get accumulo up and running on it. We have pulled down the accumulo branch
>> on github and made the packages, where do we go from here.
>>
>> Any help would be appreciated.
>>
>> Thanks in advance,
>>
>> Frank Palladino
>>
>
>
>
> --
> jay vyas
>

Re: Getting started with accumulo

[jay vyas <ja...@gmail.com>]

Hi frank.

I think sean mackrory is the main guy involved with the accumulo stuff.

- I would guess that the key thing is to grab your accumulo distribution
and confirm that it works with the exact version of hadoop you have (i.e.
hadoop version).

- Sean do you have any thoughts on how to deploy accumulo packages?  I dont
think we have any puppet recipes for them yet.

Will be interesting to  try out the accumulo branch at some point, possibly
adding smoke tests to them as well.

- In the meantime, you can also check the accumulo mailing list, maybe some
of the folks there will be interested in testing / contributing to the
accumulo packages in bigtop.




On Fri, Aug 15, 2014 at 11:45 AM, Frank Palladino <palladino.frank@gmail.com
> wrote:

> Hello Bigtop,
>
> We have a working cluster which is managed by bigtop and would like to get
> accumulo up and running on it. We have pulled down the accumulo branch on
> github and made the packages, where do we go from here.
>
> Any help would be appreciated.
>
> Thanks in advance,
>
> Frank Palladino
>



-- 
jay vyas