admin web UI

[Ran Tavory]

I'd be happy to get a management web interface on each cassandra node so i
can easily (and visually) see what's going on with the node.
Things like all the properties currently exposed by JMX and perhaps some
graphs. I'd like to simply browse to http://cassandra1:12345/monitor and
view the data and graphs.

I realize that ones JMX is exposed you can do almost everything with it, but
still I think it'd be nicer to get that kind of monitoring out of the box.
It can even be done better as a web UI, for example it can visualize how
keys are distributed in the ring, you can click links and browse the next
host, see what cassandra thinks of your rack-aware definitions etc, can be
nice for debugging and setup.

Have you seen this request in the past? What are your thoughts?

Re: admin web UI

[Sylvain Lebresne]

https://issues.apache.org/jira/browse/CASSANDRA-918

Someone seems to be working on the idea. Don't know what's the status exactly.

On Mon, May 3, 2010 at 12:49 PM, Ran Tavory  wrote:
> I'd be happy to get a management web interface on each cassandra node so i
> can easily (and visually) see what's going on with the node.
> Things like all the properties currently exposed by JMX and perhaps some
> graphs. I'd like to simply browse to http://cassandra1:12345/monitor and
> view the data and graphs.
>
> I realize that ones JMX is exposed you can do almost everything with it, but
> still I think it'd be nicer to get that kind of monitoring out of the box.
> It can even be done better as a web UI, for example it can visualize how
> keys are distributed in the ring, you can click links and browse the next
> host, see what cassandra thinks of your rack-aware definitions etc, can be
> nice for debugging and setup.
>
> Have you seen this request in the past? What are your thoughts?
>

Re: admin web UI

[Gary Dusbabek]

That was one of our ill-fated gsoc projects.  Pablo has indicated that
he is still interested in working on it, but I don't know what his
availability is.

I'm personally not in favor of the webserver-per-node approach, but I
do see its merits.  For this reason I think that it would be good that
whatever we come up with can monitor/manage a single server just was
well as a cluster.  Maybe it could treat a single node as a cluster of
one with cluster management pieces disabled.

Gary.

On Mon, May 3, 2010 at 05:58, Sylvain Lebresne  wrote:
> https://issues.apache.org/jira/browse/CASSANDRA-918
>
> Someone seems to be working on the idea. Don't know what's the status exactly.
>
> On Mon, May 3, 2010 at 12:49 PM, Ran Tavory  wrote:
>> I'd be happy to get a management web interface on each cassandra node so i
>> can easily (and visually) see what's going on with the node.
>> Things like all the properties currently exposed by JMX and perhaps some
>> graphs. I'd like to simply browse to http://cassandra1:12345/monitor and
>> view the data and graphs.
>>
>> I realize that ones JMX is exposed you can do almost everything with it, but
>> still I think it'd be nicer to get that kind of monitoring out of the box.
>> It can even be done better as a web UI, for example it can visualize how
>> keys are distributed in the ring, you can click links and browse the next
>> host, see what cassandra thinks of your rack-aware definitions etc, can be
>> nice for debugging and setup.
>>
>> Have you seen this request in the past? What are your thoughts?
>>
>

Build failed in Hudson: Cassandra #424

[Apache Hudson Server]

See 

--
[...truncated 1190 lines...]
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.db.RemoveColumnFamilyTest
[junit] Tests run: 1, Failures: 0, Errors: 0, Time elapsed: 0.603 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.db.RemoveColumnFamilyWithFlush1Test
[junit] Tests run: 1, Failures: 0, Errors: 0, Time elapsed: 0.655 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.db.RemoveColumnFamilyWithFlush2Test
[junit] Tests run: 1, Failures: 0, Errors: 0, Time elapsed: 0.669 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.db.RemoveColumnTest
[junit] Tests run: 1, Failures: 0, Errors: 0, Time elapsed: 0.694 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.db.RemoveSubColumnTest
[junit] Tests run: 1, Failures: 0, Errors: 0, Time elapsed: 0.654 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.db.RemoveSuperColumnTest
[junit] Tests run: 4, Failures: 0, Errors: 0, Time elapsed: 1.179 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.db.RowTest
[junit] Tests run: 3, Failures: 0, Errors: 0, Time elapsed: 0.472 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.db.SuperColumnTest
[junit] Tests run: 1, Failures: 0, Errors: 0, Time elapsed: 0.146 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.db.SystemTableTest
[junit] Tests run: 2, Failures: 0, Errors: 0, Time elapsed: 0.553 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.db.TableTest
[junit] Tests run: 10, Failures: 0, Errors: 0, Time elapsed: 5.097 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.db.TimeSortTest
[junit] Tests run: 2, Failures: 0, Errors: 0, Time elapsed: 5.447 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.db.commitlog.CommitLogHeaderTest
[junit] Tests run: 6, Failures: 0, Errors: 0, Time elapsed: 0.481 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.db.marshal.TypeCompareTest
[junit] Tests run: 4, Failures: 0, Errors: 0, Time elapsed: 0.158 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.dht.BootStrapperTest
[junit] Tests run: 3, Failures: 0, Errors: 0, Time elapsed: 1.634 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.dht.BoundsTest
[junit] Tests run: 4, Failures: 0, Errors: 0, Time elapsed: 0.479 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.dht.ByteOrderedPartitionerTest
[junit] Tests run: 6, Failures: 0, Errors: 0, Time elapsed: 10.322 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: 
org.apache.cassandra.dht.CollatingOrderPreservingPartitionerTest
[junit] Tests run: 8, Failures: 0, Errors: 0, Time elapsed: 12.149 sec
[junit] 
[junit] Cobertura: Loaded information on 786 classes.
[junit] Cobertura: Saved information on 786 classes.
[junit] Testsuite: org.apache.cassandra.dht.OrderPreservingPartitionerTest
[junit] Tests run: 7, Failures: 0, Errors: 0, Time 

Hudson build is back to normal : Cassandra #425

[Apache Hudson Server]

See 

我在 Facebook 找你呢

[俊毅吳]

我開設了一個 Facebook 個人檔案，內有我的相片、影片及活動。我想將你列為我在 Facebook 上的朋友，這樣你就可以看到我的個人檔案。你首先要加入 
Facebook 。然後你也可以設立你自己的個人檔案。

謝謝，
吳

欲註冊 Facebook，請點取下列連結：
http://www.facebook.com/p.php?i=11049690791&k=Z6E3Y6W63XWGXKEJPB63QTYUZRAFYT2NUWGSDKQ1UQ&r



Already have an account? Add this email address to your account 
http://www.facebook.com/n/?merge_accounts.php&e=dev%40cassandra.apache.org&c=4910bd82496234cce2546e9949bd8700.俊毅吳邀請
 dev@cassandra.apache.org 加入Facebook。 若是您不想再收到 Facebook 寄來的此類信件，請按下列的連結以取消訂閱。
http://www.facebook.com/o.php?k=6e5ad9&u=10869725694&mid=2472498G5af3445135feG0G8
Facebook 公司地址為 1601 S. California Ave., Palo Alto, CA 94304 。

Re: admin web UI

[Eric Evans]

On Mon, 2010-05-03 at 12:58 +0200, Sylvain Lebresne wrote:
> https://issues.apache.org/jira/browse/CASSANDRA-918
> 
> Someone seems to be working on the idea. Don't know what's the status
> exactly. 

#918 is to create a centralized, browser-based, management station,
(FWIW, I think this is The Way).

Earlier versions of Cassandra had an embedded web server on each node
that displayed a few properties, including a non-graphical
representation (a table) of the ring. We removed this webui because as
time wore on, more and more people came around to the idea that it was
better to put the time into JMX instrumentation and leave the job of
monitoring/management to applications and developers better suited to
that. Technically speaking, we removed it after this overall attitude
resulted in lots of excellent work being put into JMX, while the webui
suffered from chronic lack of maintenance.

A webui has a certain gee-whiz factor, but when running a real cluster
of any size, it's probably the least practical way to go about it IMO.

-- 
Eric Evans
eev...@rackspace.com

Re: admin web UI

[Sylvain Lebresne]

For the record, I was just pointing #918 for the web admin gui part but
I totally agree and also think that a centralized management thingy is the
way to go.

On Mon, May 3, 2010 at 5:38 PM, Eric Evans  wrote:
> On Mon, 2010-05-03 at 12:58 +0200, Sylvain Lebresne wrote:
>> https://issues.apache.org/jira/browse/CASSANDRA-918
>>
>> Someone seems to be working on the idea. Don't know what's the status
>> exactly.
>
> #918 is to create a centralized, browser-based, management station,
> (FWIW, I think this is The Way).
>
> Earlier versions of Cassandra had an embedded web server on each node
> that displayed a few properties, including a non-graphical
> representation (a table) of the ring. We removed this webui because as
> time wore on, more and more people came around to the idea that it was
> better to put the time into JMX instrumentation and leave the job of
> monitoring/management to applications and developers better suited to
> that. Technically speaking, we removed it after this overall attitude
> resulted in lots of excellent work being put into JMX, while the webui
> suffered from chronic lack of maintenance.
>
> A webui has a certain gee-whiz factor, but when running a real cluster
> of any size, it's probably the least practical way to go about it IMO.
>
> --
> Eric Evans
> eev...@rackspace.com
>
>

Skip large size (Configurable) SSTable in minor or/and major compaction

[Schubert Zhang]

We make a patch to 0.6 branch and 0.6.1 for this feature.

https://issues.apache.org/jira/browse/CASSANDRA-1041

Re: admin web UI

[Ran Tavory]

How about the following compromise:
Add a simple web server to each node with only one simple servlet that
simply spits out all JMX stats on one page. Not fancy, no graphs, simply the
same values you can get from jconsole, but on a web page.
To me it seems like a fair tradeoff b/w maintenance and easier out of the
box management.  Shooting up jconsole for each server is cumbersome, at
least in the environment I work in (firewalls, high latency etc) so a web
interface can be nice.

Regardless of this, I agree a fancy external monitoring tool is in place,
https://issues.apache.org/jira/browse/CASSANDRA-918

On Mon, May 3, 2010 at 6:46 PM, Sylvain Lebresne  wrote:

> For the record, I was just pointing #918 for the web admin gui part but
> I totally agree and also think that a centralized management thingy is the
> way to go.
>
> On Mon, May 3, 2010 at 5:38 PM, Eric Evans  wrote:
> > On Mon, 2010-05-03 at 12:58 +0200, Sylvain Lebresne wrote:
> >> https://issues.apache.org/jira/browse/CASSANDRA-918
> >>
> >> Someone seems to be working on the idea. Don't know what's the status
> >> exactly.
> >
> > #918 is to create a centralized, browser-based, management station,
> > (FWIW, I think this is The Way).
> >
> > Earlier versions of Cassandra had an embedded web server on each node
> > that displayed a few properties, including a non-graphical
> > representation (a table) of the ring. We removed this webui because as
> > time wore on, more and more people came around to the idea that it was
> > better to put the time into JMX instrumentation and leave the job of
> > monitoring/management to applications and developers better suited to
> > that. Technically speaking, we removed it after this overall attitude
> > resulted in lots of excellent work being put into JMX, while the webui
> > suffered from chronic lack of maintenance.
> >
> > A webui has a certain gee-whiz factor, but when running a real cluster
> > of any size, it's probably the least practical way to go about it IMO.
> >
> > --
> > Eric Evans
> > eev...@rackspace.com
> >
> >
>