On Wednesday 18 June 2008, Dennis Kubes wrote:
> was to
> may the Nutch architecture itself more flexible. This is what I have
> been terming Nutch 2 and what I have been currently working on.
Sounds interesting. Can you point me to any further information or discussion
threads on Nutch 2?
Isab
Michael Bieniosek reports I'd mangled release candidate 2.
Here is release candidate 3:
http://people.apache.org/~stack/hbase-0.1.3-candidate-3/
Please take it for a spin. Voting (again) closes Friday, June 20th.
This release only runs on hadoop 0.16.x.
Yours,
HBase Team
stack wrote:
Here
Clint Morgan wrote:
I have a local cluster running, and its logging to
/log_X.X.X.X_1213228101021_60020/
Then I kill both master and regionserver, and restart. Looking through
the logs I don't see anything about trying to recover from this hlog,
it just creates a new hlog alongside the existing
Here's a new 0.1.3 release candidate:
http://people.apache.org/~stack/hbase-0.1.3-candidate-2/
Please try it out. Voting closes Friday, June 20th.
(Candidate 1 was squashed because a couple of more fixes came in:
HBASE-680, HBASE-684, and HBASE-686)
Thanks,
The HBase Team
We have discussed it but not implemented it. A previous step before
implementing interfaces to use HBase for current Nutch databases was to
may the Nutch architecture itself more flexible. This is what I have
been terming Nutch 2 and what I have been currently working on.
Dennis
Marcus Hero
I ain't sure how it'd work -- a callback that regionserver-side scanner
can pull on to send the state-laden filter back to the client for
passing to next region? -- but I like this idea. I added hbase-695.
St.Ack
Clint Morgan wrote:
Ah, good point. I missed that.
I guess to make this work we
Ah, good point. I missed that.
I guess to make this work we need to add an RCP method on regionserver
to get the current filter, grab it clientside when done with a region,
and pass it back to the next region to be scanned (rather than a fresh
filter as currently).
2008/6/18 stack <[EMAIL PROTECT
Enable DEBUG. Then look at master and regionserver log. My guess is
that there is something up on the regionserver that is preventing
successful deploy of the region assigned by the master.
St.Ack
Krzysztof Szlapinski wrote:
hi all,
I'm just in the middle of a process of creation of more th
I was going to suggest filters but was wondering what happens when you
cross regions or cross regionservers? What happens if regionserver A
only has half of the fifth page and you need to go to another server to
get the rest? Its filters will know nought of regionserver A's state?
Thanks,
St
Why the aversion to filters? Thats how we solve this problem, I have a
simple SkipRowFilter that I wrote that does exactly this...
-clint
2008/6/18 Krzysztof Gałęcki <[EMAIL PROTECTED]>:
> I can't cache items on List because of two main reasons:
>
> 1. I have to many items to cache them in memory
I can't cache items on List because of two main reasons:
1. I have to many items to cache them in memory
2. I would also like to use non GUID-based rowIds. In that situation we are
talking about inserting items, not just adding them to the end of table
(list).
I think there should be some method
Krzych:
What J-D says -- an application-side cache associated with user session -- or
stash startrows into user session? For example, on first fetch, cut up the
first hundred or so results into pages and save the page startrows in user
session. When user comes back and asks for a 3rd page, s
hi all,
I'm just in the middle of a process of creation of more than 30 empty
tables in a row on a cluster of 3 machines
I've just ran a batch script in the jruby shell and I'm observing what
happens on the different Region Servers
I can see that HMaster during this process of creation keeps re
Hi krzych,
If you really need help fast, come see us on IRC. Details are on the
website.
My solution for your problem would be to cache the results in a List instead
of always scanning the table.
jdcryans
On Wed, Jun 18, 2008 at 11:20 AM, Krzysztof Gałęcki
<[EMAIL PROTECTED]> wrote:
> Sorry fo
Sorry for spam, but looks like I send this message as a replay to some other
post, so one more time:
Hi
I have a problem with scanner on HTable object: I have a table with
GUID-based rowId. I would like to display items from this table on pages
(for example 10 items on each page). I'm cr
On Wed, Jun 18, 2008 at 3:21 PM, yunyi.x <[EMAIL PROTECTED]> wrote:
>
> can i save lots of data into one cell with different column tag names?
> like:
>
> |rowKey | TimeStamp | Column |
> |-|
>
can i save lots of data into one cell with different column tag names?
like:
|rowKey | TimeStamp | Column |
|-|
| key1 | 234255645 |colFamily:tag1 | value1 |
|-
can i save lots of data into one cell with different column tag names?
like:
|rowKey | TimeStamp | Column |
|-|
| key1 | 234255645 |colFamily:tag1 | value1 |
|-
18 matches
Mail list logo
