[jira] Created: (NUTCH-395) Increase fetching speed

classic Classic list List threaded Threaded
17 messages Options
Reply | Threaded
Open this post in threaded view
|

[jira] Created: (NUTCH-395) Increase fetching speed

Nick Burch (Jira)
Increase fetching speed
-----------------------

                 Key: NUTCH-395
                 URL: http://issues.apache.org/jira/browse/NUTCH-395
             Project: Nutch
          Issue Type: Improvement
          Components: fetcher
    Affects Versions: 0.8.1
            Reporter: Sami Siren
         Assigned To: Sami Siren


There have been some discussion on nutch mailing lists about fetcher being slow, this patch tried to address that. the patch is just a quich hack and needs some cleaning up, it also currently applies to 0.8 branch and not trunk and it has also not been tested in large. What it changes?

Metadata - the original metadata uses spellchecking, new version does not (a decorator is provided that can do it and it should perhaps be used where http headers are handled but in most of the cases the functionality is not required)

Reading/writing various data structures - patch tries to do io more efficiently see the patch for details.

Initial benchmark:

A small benchmark was done to measure the performance of changes with a script that basically does the following:
-inject a list of urls into a fresh crawldb
-create fetchlist (10k urls pointing to local filesystem)
-fetch
-updatedb

original code from 0.8-branch:
real    10m51.907s
user    10m9.914s
sys     0m21.285s

after applying the patch
real    4m15.313s
user    3m42.598s
sys     0m18.485s



--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       
Reply | Threaded
Open this post in threaded view
|

[jira] Updated: (NUTCH-395) Increase fetching speed

Nick Burch (Jira)
     [ http://issues.apache.org/jira/browse/NUTCH-395?page=all ]

Sami Siren updated NUTCH-395:
-----------------------------

    Attachment: nutch-0.8-performance.txt

a rough patch for testing purposes

> Increase fetching speed
> -----------------------
>
>                 Key: NUTCH-395
>                 URL: http://issues.apache.org/jira/browse/NUTCH-395
>             Project: Nutch
>          Issue Type: Improvement
>          Components: fetcher
>    Affects Versions: 0.8.1
>            Reporter: Sami Siren
>         Assigned To: Sami Siren
>         Attachments: nutch-0.8-performance.txt
>
>
> There have been some discussion on nutch mailing lists about fetcher being slow, this patch tried to address that. the patch is just a quich hack and needs some cleaning up, it also currently applies to 0.8 branch and not trunk and it has also not been tested in large. What it changes?
> Metadata - the original metadata uses spellchecking, new version does not (a decorator is provided that can do it and it should perhaps be used where http headers are handled but in most of the cases the functionality is not required)
> Reading/writing various data structures - patch tries to do io more efficiently see the patch for details.
> Initial benchmark:
> A small benchmark was done to measure the performance of changes with a script that basically does the following:
> -inject a list of urls into a fresh crawldb
> -create fetchlist (10k urls pointing to local filesystem)
> -fetch
> -updatedb
> original code from 0.8-branch:
> real    10m51.907s
> user    10m9.914s
> sys     0m21.285s
> after applying the patch
> real    4m15.313s
> user    3m42.598s
> sys     0m18.485s

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       
Reply | Threaded
Open this post in threaded view
|

[jira] Commented: (NUTCH-395) Increase fetching speed

Nick Burch (Jira)
In reply to this post by Nick Burch (Jira)
    [ http://issues.apache.org/jira/browse/NUTCH-395?page=comments#action_12445532 ]
           
Andrzej Bialecki  commented on NUTCH-395:
-----------------------------------------

I have several comments to this patch:

* have you measured what made the biggest impact on performance - changes to Metadata, or changes to IO in FetcherOutput?

* I think it's a good idea to separate two concerns with PlainMetadata / MetadataSpellChecker. Since the latter is a subclass I think it would be more appropriate to name it SpellCheckedMetadata.

* I'd also argue for keeping the name Metadata and just replace the body of the class with PlainMetadata implementation - this way we could avoid changing the API in so many places; for compatibility we could just bump the version number in Metadata. We could then avoid also changes to version id-s of other classes that rely on Metadata, such as Content, ParseData et al.

* new Metadata / SpellCheckedMetadata need JUnit tests - this is important, because many other classes rely on proper working of these classes.

* Fetcher.VoidReducer is not needed - I'm guessing you wanted to use it just for logging.

* please observe formatting rules, especially whitespace rules - this patch doesn't follow them.

> Increase fetching speed
> -----------------------
>
>                 Key: NUTCH-395
>                 URL: http://issues.apache.org/jira/browse/NUTCH-395
>             Project: Nutch
>          Issue Type: Improvement
>          Components: fetcher
>    Affects Versions: 0.8.1
>            Reporter: Sami Siren
>         Assigned To: Sami Siren
>         Attachments: nutch-0.8-performance.txt
>
>
> There have been some discussion on nutch mailing lists about fetcher being slow, this patch tried to address that. the patch is just a quich hack and needs some cleaning up, it also currently applies to 0.8 branch and not trunk and it has also not been tested in large. What it changes?
> Metadata - the original metadata uses spellchecking, new version does not (a decorator is provided that can do it and it should perhaps be used where http headers are handled but in most of the cases the functionality is not required)
> Reading/writing various data structures - patch tries to do io more efficiently see the patch for details.
> Initial benchmark:
> A small benchmark was done to measure the performance of changes with a script that basically does the following:
> -inject a list of urls into a fresh crawldb
> -create fetchlist (10k urls pointing to local filesystem)
> -fetch
> -updatedb
> original code from 0.8-branch:
> real    10m51.907s
> user    10m9.914s
> sys     0m21.285s
> after applying the patch
> real    4m15.313s
> user    3m42.598s
> sys     0m18.485s

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       
Reply | Threaded
Open this post in threaded view
|

[jira] Commented: (NUTCH-395) Increase fetching speed

Nick Burch (Jira)
In reply to this post by Nick Burch (Jira)
    [ http://issues.apache.org/jira/browse/NUTCH-395?page=comments#action_12445956 ]
           
Sami Siren commented on NUTCH-395:
----------------------------------

>have you measured what made the biggest impact on performance - changes to Metadata, or
>changes to IO in FetcherOutput?
did not have time yet, I would quess that IO changes make most signifigant part.

>I'd also argue for keeping the name Metadata and just replace the body of the class with PlainMetadata
>implementation - this way we could avoid changing the API in so many places; for compatibility we could
>just bump the version number in Metadata. We could then avoid also changes to version id-s of other
>classes that rely on Metadata, such as Content, ParseData et al.

The api for new metadata is exactly the same, but the functionality changed so I decided to make a new class totally, but Yes I agree here, It's much more clean to replace the guts of Metadata class.

>new Metadata / SpellCheckedMetadata need JUnit tests - this is important, because many other classes rely
>on proper working of these classes.
sure, there was supposed to be some allready in the patch but I just forgot to svn add them.

Now that I remember, there was one more odd thing in current implementation: the max number of links was not enforced when writing outlinks only when reading them, I am planning to change this also so the number of links is enforced on write.

>Fetcher.VoidReducer is not needed - I'm guessing you wanted to use it just for logging.
true

>please observe formatting rules, especially whitespace rules - this patch doesn't follow them.

will do, as I said this was not meant to be a demonstration of nice formatting or java coding, just wanted to throw out the
findings for people to try them out. I'll start to work on a new version against trunk  and will do it with more focusused mindset :)

> Increase fetching speed
> -----------------------
>
>                 Key: NUTCH-395
>                 URL: http://issues.apache.org/jira/browse/NUTCH-395
>             Project: Nutch
>          Issue Type: Improvement
>          Components: fetcher
>    Affects Versions: 0.8.1
>            Reporter: Sami Siren
>         Assigned To: Sami Siren
>         Attachments: nutch-0.8-performance.txt
>
>
> There have been some discussion on nutch mailing lists about fetcher being slow, this patch tried to address that. the patch is just a quich hack and needs some cleaning up, it also currently applies to 0.8 branch and not trunk and it has also not been tested in large. What it changes?
> Metadata - the original metadata uses spellchecking, new version does not (a decorator is provided that can do it and it should perhaps be used where http headers are handled but in most of the cases the functionality is not required)
> Reading/writing various data structures - patch tries to do io more efficiently see the patch for details.
> Initial benchmark:
> A small benchmark was done to measure the performance of changes with a script that basically does the following:
> -inject a list of urls into a fresh crawldb
> -create fetchlist (10k urls pointing to local filesystem)
> -fetch
> -updatedb
> original code from 0.8-branch:
> real    10m51.907s
> user    10m9.914s
> sys     0m21.285s
> after applying the patch
> real    4m15.313s
> user    3m42.598s
> sys     0m18.485s

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       
Reply | Threaded
Open this post in threaded view
|

[jira] Commented: (NUTCH-395) Increase fetching speed

Nick Burch (Jira)
In reply to this post by Nick Burch (Jira)
    [ http://issues.apache.org/jira/browse/NUTCH-395?page=comments#action_12445994 ]
           
Andrzej Bialecki  commented on NUTCH-395:
-----------------------------------------

> Now that I remember, there was one more odd thing in current implementation: the max number
> of links was not enforced when writing outlinks only when reading them, I am planning to change
> this also so the number of links is enforced on write.

AFAIK this was done on purpose, to facilitate processing of existing data created with different settings. I.e. if someone created a segment with high max  # of outlinks, you should still be able to read it and process all outlinks. If you enforce the max # during reading you won't be able to process this data.

> Increase fetching speed
> -----------------------
>
>                 Key: NUTCH-395
>                 URL: http://issues.apache.org/jira/browse/NUTCH-395
>             Project: Nutch
>          Issue Type: Improvement
>          Components: fetcher
>    Affects Versions: 0.8.1
>            Reporter: Sami Siren
>         Assigned To: Sami Siren
>         Attachments: nutch-0.8-performance.txt
>
>
> There have been some discussion on nutch mailing lists about fetcher being slow, this patch tried to address that. the patch is just a quich hack and needs some cleaning up, it also currently applies to 0.8 branch and not trunk and it has also not been tested in large. What it changes?
> Metadata - the original metadata uses spellchecking, new version does not (a decorator is provided that can do it and it should perhaps be used where http headers are handled but in most of the cases the functionality is not required)
> Reading/writing various data structures - patch tries to do io more efficiently see the patch for details.
> Initial benchmark:
> A small benchmark was done to measure the performance of changes with a script that basically does the following:
> -inject a list of urls into a fresh crawldb
> -create fetchlist (10k urls pointing to local filesystem)
> -fetch
> -updatedb
> original code from 0.8-branch:
> real    10m51.907s
> user    10m9.914s
> sys     0m21.285s
> after applying the patch
> real    4m15.313s
> user    3m42.598s
> sys     0m18.485s

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       
Reply | Threaded
Open this post in threaded view
|

[jira] Commented: (NUTCH-395) Increase fetching speed

Nick Burch (Jira)
In reply to this post by Nick Burch (Jira)
    [ http://issues.apache.org/jira/browse/NUTCH-395?page=comments#action_12445999 ]
           
Sami Siren commented on NUTCH-395:
----------------------------------

> settings. I.e. if someone created a segment with high max # of outlinks, you should still be able
> to read it and process all outlinks. If you enforce the max # during reading you won't be able
> to process this data.

Yes i agree, but IMO we should also not store more than configured max # of links, now it seems we
store em all (or am i just not seeing it?).

> Increase fetching speed
> -----------------------
>
>                 Key: NUTCH-395
>                 URL: http://issues.apache.org/jira/browse/NUTCH-395
>             Project: Nutch
>          Issue Type: Improvement
>          Components: fetcher
>    Affects Versions: 0.8.1
>            Reporter: Sami Siren
>         Assigned To: Sami Siren
>         Attachments: nutch-0.8-performance.txt
>
>
> There have been some discussion on nutch mailing lists about fetcher being slow, this patch tried to address that. the patch is just a quich hack and needs some cleaning up, it also currently applies to 0.8 branch and not trunk and it has also not been tested in large. What it changes?
> Metadata - the original metadata uses spellchecking, new version does not (a decorator is provided that can do it and it should perhaps be used where http headers are handled but in most of the cases the functionality is not required)
> Reading/writing various data structures - patch tries to do io more efficiently see the patch for details.
> Initial benchmark:
> A small benchmark was done to measure the performance of changes with a script that basically does the following:
> -inject a list of urls into a fresh crawldb
> -create fetchlist (10k urls pointing to local filesystem)
> -fetch
> -updatedb
> original code from 0.8-branch:
> real    10m51.907s
> user    10m9.914s
> sys     0m21.285s
> after applying the patch
> real    4m15.313s
> user    3m42.598s
> sys     0m18.485s

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       
Reply | Threaded
Open this post in threaded view
|

[jira] Commented: (NUTCH-395) Increase fetching speed

Nick Burch (Jira)
In reply to this post by Nick Burch (Jira)
    [ http://issues.apache.org/jira/browse/NUTCH-395?page=comments#action_12448795 ]
           
Sami Siren commented on NUTCH-395:
----------------------------------

>>have you measured what made the biggest impact on performance - changes to Metadata, or
>>changes to IO in FetcherOutput?
>did not have time yet, I would quess that IO changes make most signifigant part.

After more digging my initial guess might not have been correct. By not touching IO at all
I am able to get same improvement changing the trunk when comparing to nightly builds as
I reported before on 0.8 branch.

This is good, because we don't need to change file formats at all.



> Increase fetching speed
> -----------------------
>
>                 Key: NUTCH-395
>                 URL: http://issues.apache.org/jira/browse/NUTCH-395
>             Project: Nutch
>          Issue Type: Improvement
>          Components: fetcher
>    Affects Versions: 0.8.1
>            Reporter: Sami Siren
>         Assigned To: Sami Siren
>         Attachments: nutch-0.8-performance.txt
>
>
> There have been some discussion on nutch mailing lists about fetcher being slow, this patch tried to address that. the patch is just a quich hack and needs some cleaning up, it also currently applies to 0.8 branch and not trunk and it has also not been tested in large. What it changes?
> Metadata - the original metadata uses spellchecking, new version does not (a decorator is provided that can do it and it should perhaps be used where http headers are handled but in most of the cases the functionality is not required)
> Reading/writing various data structures - patch tries to do io more efficiently see the patch for details.
> Initial benchmark:
> A small benchmark was done to measure the performance of changes with a script that basically does the following:
> -inject a list of urls into a fresh crawldb
> -create fetchlist (10k urls pointing to local filesystem)
> -fetch
> -updatedb
> original code from 0.8-branch:
> real    10m51.907s
> user    10m9.914s
> sys     0m21.285s
> after applying the patch
> real    4m15.313s
> user    3m42.598s
> sys     0m18.485s

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       
Reply | Threaded
Open this post in threaded view
|

[jira] Updated: (NUTCH-395) Increase fetching speed

Nick Burch (Jira)
In reply to this post by Nick Burch (Jira)
     [ http://issues.apache.org/jira/browse/NUTCH-395?page=all ]

Sami Siren updated NUTCH-395:
-----------------------------

    Attachment: NUTCH-395-trunk-metadata-only.patch

Here's a first stab at svn trunk version of nutch that just optimizes the use of metadata and splits it into two functionally distict pieces one for plain metadata and one for spellchecking over the keys of metadata.

There's propably still room for optimization on both the metadata and IO side also.

The same local filesystem fetching bench was run as earlier, this time on trunk version. Even if the benchmark was run witl file:// urls it should affect other protocols also specifically because it seems to cut down the time needed for reduce phase quite aggressively.

I would also recommend adding some kind of base benchmark for crawling operations to nutch so we don't kill the performance (again and again) at some point.

from svn trunk
----------------------
real    10m43.527s
user    10m11.210s
sys     0m21.837s

fetch breakdown:
5 min 19 sec effective fetching
7 sec sort
4 min 30 sec reduce > reduce


patched version
----------------------
real    4m53.742s
user    4m21.340s
sys     0m19.045s

fetch breakdown:
3 min 36 sec effective fetching
8 sec sort
27 sec reduce > reduce



> Increase fetching speed
> -----------------------
>
>                 Key: NUTCH-395
>                 URL: http://issues.apache.org/jira/browse/NUTCH-395
>             Project: Nutch
>          Issue Type: Improvement
>          Components: fetcher
>    Affects Versions: 0.9.0, 0.8.1
>            Reporter: Sami Siren
>         Assigned To: Sami Siren
>         Attachments: nutch-0.8-performance.txt, NUTCH-395-trunk-metadata-only.patch
>
>
> There have been some discussion on nutch mailing lists about fetcher being slow, this patch tried to address that. the patch is just a quich hack and needs some cleaning up, it also currently applies to 0.8 branch and not trunk and it has also not been tested in large. What it changes?
> Metadata - the original metadata uses spellchecking, new version does not (a decorator is provided that can do it and it should perhaps be used where http headers are handled but in most of the cases the functionality is not required)
> Reading/writing various data structures - patch tries to do io more efficiently see the patch for details.
> Initial benchmark:
> A small benchmark was done to measure the performance of changes with a script that basically does the following:
> -inject a list of urls into a fresh crawldb
> -create fetchlist (10k urls pointing to local filesystem)
> -fetch
> -updatedb
> original code from 0.8-branch:
> real    10m51.907s
> user    10m9.914s
> sys     0m21.285s
> after applying the patch
> real    4m15.313s
> user    3m42.598s
> sys     0m18.485s

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       
Reply | Threaded
Open this post in threaded view
|

[jira] Updated: (NUTCH-395) Increase fetching speed

Nick Burch (Jira)
In reply to this post by Nick Burch (Jira)
     [ http://issues.apache.org/jira/browse/NUTCH-395?page=all ]

Sami Siren updated NUTCH-395:
-----------------------------

    Affects Version/s: 0.9.0

> Increase fetching speed
> -----------------------
>
>                 Key: NUTCH-395
>                 URL: http://issues.apache.org/jira/browse/NUTCH-395
>             Project: Nutch
>          Issue Type: Improvement
>          Components: fetcher
>    Affects Versions: 0.9.0, 0.8.1
>            Reporter: Sami Siren
>         Assigned To: Sami Siren
>         Attachments: nutch-0.8-performance.txt, NUTCH-395-trunk-metadata-only.patch
>
>
> There have been some discussion on nutch mailing lists about fetcher being slow, this patch tried to address that. the patch is just a quich hack and needs some cleaning up, it also currently applies to 0.8 branch and not trunk and it has also not been tested in large. What it changes?
> Metadata - the original metadata uses spellchecking, new version does not (a decorator is provided that can do it and it should perhaps be used where http headers are handled but in most of the cases the functionality is not required)
> Reading/writing various data structures - patch tries to do io more efficiently see the patch for details.
> Initial benchmark:
> A small benchmark was done to measure the performance of changes with a script that basically does the following:
> -inject a list of urls into a fresh crawldb
> -create fetchlist (10k urls pointing to local filesystem)
> -fetch
> -updatedb
> original code from 0.8-branch:
> real    10m51.907s
> user    10m9.914s
> sys     0m21.285s
> after applying the patch
> real    4m15.313s
> user    3m42.598s
> sys     0m18.485s

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       
Reply | Threaded
Open this post in threaded view
|

[jira] Updated: (NUTCH-395) Increase fetching speed

Nick Burch (Jira)
In reply to this post by Nick Burch (Jira)
     [ http://issues.apache.org/jira/browse/NUTCH-395?page=all ]

Sami Siren updated NUTCH-395:
-----------------------------

    Attachment: NUTCH-395-trunk-metadata-only-2.patch

Additional change to Content cuts down time needed in effective fetching. Now seeing speeds like 45 pages/sec also on http.

real    4m24.126s
user    3m53.835s
sys     0m18.681s

3 min 10 sec effective fetching
6 sec sorting
27 sec  reduce > reduce

> Increase fetching speed
> -----------------------
>
>                 Key: NUTCH-395
>                 URL: http://issues.apache.org/jira/browse/NUTCH-395
>             Project: Nutch
>          Issue Type: Improvement
>          Components: fetcher
>    Affects Versions: 0.9.0, 0.8.1
>            Reporter: Sami Siren
>         Assigned To: Sami Siren
>         Attachments: nutch-0.8-performance.txt, NUTCH-395-trunk-metadata-only-2.patch, NUTCH-395-trunk-metadata-only.patch
>
>
> There have been some discussion on nutch mailing lists about fetcher being slow, this patch tried to address that. the patch is just a quich hack and needs some cleaning up, it also currently applies to 0.8 branch and not trunk and it has also not been tested in large. What it changes?
> Metadata - the original metadata uses spellchecking, new version does not (a decorator is provided that can do it and it should perhaps be used where http headers are handled but in most of the cases the functionality is not required)
> Reading/writing various data structures - patch tries to do io more efficiently see the patch for details.
> Initial benchmark:
> A small benchmark was done to measure the performance of changes with a script that basically does the following:
> -inject a list of urls into a fresh crawldb
> -create fetchlist (10k urls pointing to local filesystem)
> -fetch
> -updatedb
> original code from 0.8-branch:
> real    10m51.907s
> user    10m9.914s
> sys     0m21.285s
> after applying the patch
> real    4m15.313s
> user    3m42.598s
> sys     0m18.485s

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       
Reply | Threaded
Open this post in threaded view
|

[jira] Commented: (NUTCH-395) Increase fetching speed

Nick Burch (Jira)
In reply to this post by Nick Burch (Jira)
    [ http://issues.apache.org/jira/browse/NUTCH-395?page=comments#action_12449292 ]
           
Andrzej Bialecki  commented on NUTCH-395:
-----------------------------------------

+1 - this patch looks good to me - if you could just fix the whitespace issues prior to committing, so that it conforms to the coding style ...

> Increase fetching speed
> -----------------------
>
>                 Key: NUTCH-395
>                 URL: http://issues.apache.org/jira/browse/NUTCH-395
>             Project: Nutch
>          Issue Type: Improvement
>          Components: fetcher
>    Affects Versions: 0.9.0, 0.8.1
>            Reporter: Sami Siren
>         Assigned To: Sami Siren
>         Attachments: nutch-0.8-performance.txt, NUTCH-395-trunk-metadata-only-2.patch, NUTCH-395-trunk-metadata-only.patch
>
>
> There have been some discussion on nutch mailing lists about fetcher being slow, this patch tried to address that. the patch is just a quich hack and needs some cleaning up, it also currently applies to 0.8 branch and not trunk and it has also not been tested in large. What it changes?
> Metadata - the original metadata uses spellchecking, new version does not (a decorator is provided that can do it and it should perhaps be used where http headers are handled but in most of the cases the functionality is not required)
> Reading/writing various data structures - patch tries to do io more efficiently see the patch for details.
> Initial benchmark:
> A small benchmark was done to measure the performance of changes with a script that basically does the following:
> -inject a list of urls into a fresh crawldb
> -create fetchlist (10k urls pointing to local filesystem)
> -fetch
> -updatedb
> original code from 0.8-branch:
> real    10m51.907s
> user    10m9.914s
> sys     0m21.285s
> after applying the patch
> real    4m15.313s
> user    3m42.598s
> sys     0m18.485s

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       
Reply | Threaded
Open this post in threaded view
|

[jira] Resolved: (NUTCH-395) Increase fetching speed

Nick Burch (Jira)
In reply to this post by Nick Burch (Jira)
     [ http://issues.apache.org/jira/browse/NUTCH-395?page=all ]

Sami Siren resolved NUTCH-395.
------------------------------

    Fix Version/s: 0.9.0
       Resolution: Fixed

applied to trunk with some additional whitespace changes.

> Increase fetching speed
> -----------------------
>
>                 Key: NUTCH-395
>                 URL: http://issues.apache.org/jira/browse/NUTCH-395
>             Project: Nutch
>          Issue Type: Improvement
>          Components: fetcher
>    Affects Versions: 0.8.1, 0.9.0
>            Reporter: Sami Siren
>         Assigned To: Sami Siren
>             Fix For: 0.9.0
>
>         Attachments: nutch-0.8-performance.txt, NUTCH-395-trunk-metadata-only-2.patch, NUTCH-395-trunk-metadata-only.patch
>
>
> There have been some discussion on nutch mailing lists about fetcher being slow, this patch tried to address that. the patch is just a quich hack and needs some cleaning up, it also currently applies to 0.8 branch and not trunk and it has also not been tested in large. What it changes?
> Metadata - the original metadata uses spellchecking, new version does not (a decorator is provided that can do it and it should perhaps be used where http headers are handled but in most of the cases the functionality is not required)
> Reading/writing various data structures - patch tries to do io more efficiently see the patch for details.
> Initial benchmark:
> A small benchmark was done to measure the performance of changes with a script that basically does the following:
> -inject a list of urls into a fresh crawldb
> -create fetchlist (10k urls pointing to local filesystem)
> -fetch
> -updatedb
> original code from 0.8-branch:
> real    10m51.907s
> user    10m9.914s
> sys     0m21.285s
> after applying the patch
> real    4m15.313s
> user    3m42.598s
> sys     0m18.485s

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       
Reply | Threaded
Open this post in threaded view
|

Re: [jira] Resolved: (NUTCH-395) Increase fetching speed

AJ Chen-2
Sami,
Thanks for resolving this serious issue.  I just updated my code from trunk
and plan to test fetch speed. But ,there is a runtime error related to
switching from UTF8 to Text. Since the error is from hadoop, how do I fix
it?

java.lang.ClassCastException: org.apache.hadoop.io.UTF8
    at org.apache.nutch.crawl.Generato r$Selector.map(Generator.java:108)
    at org.apache.hadoop.mapred.MapRunner.run(MapRunner.java:46)
    at org.apache.hadoop.mapred.MapTask.run(MapTask.java:213)
    at org.apache.hadoop.mapred.LocalJobRunner$Job.run(LocalJobRunner.java
:105)

Thanks,
AJ


On 11/13/06, Sami Siren (JIRA) <[hidden email]> wrote:

>
>      [ http://issues.apache.org/jira/browse/NUTCH-395?page=all ]
>
> Sami Siren resolved NUTCH-395.
> ------------------------------
>
>     Fix Version/s: 0.9.0
>        Resolution: Fixed
>
> applied to trunk with some additional whitespace changes.
>
> > Increase fetching speed
> > -----------------------
> >
> >                 Key: NUTCH-395
> >                 URL: http://issues.apache.org/jira/browse/NUTCH-395
> >             Project: Nutch
> >          Issue Type: Improvement
> >          Components: fetcher
> >    Affects Versions: 0.8.1, 0.9.0
> >            Reporter: Sami Siren
> >         Assigned To: Sami Siren
> >             Fix For: 0.9.0
> >
> >         Attachments: nutch-0.8-performance.txt,
> NUTCH-395-trunk-metadata-only-2.patch, NUTCH-395-trunk-metadata-only.patch
> >
> >
> > There have been some discussion on nutch mailing lists about fetcher
> being slow, this patch tried to address that. the patch is just a quich hack
> and needs some cleaning up, it also currently applies to 0.8 branch and
> not trunk and it has also not been tested in large. What it changes?
> > Metadata - the original metadata uses spellchecking, new version does
> not (a decorator is provided that can do it and it should perhaps be used
> where http headers are handled but in most of the cases the functionality is
> not required)
> > Reading/writing various data structures - patch tries to do io more
> efficiently see the patch for details.
> > Initial benchmark:
> > A small benchmark was done to measure the performance of changes with a
> script that basically does the following:
> > -inject a list of urls into a fresh crawldb
> > -create fetchlist (10k urls pointing to local filesystem)
> > -fetch
> > -updatedb
> > original code from 0.8-branch:
> > real    10m51.907s
> > user    10m9.914s
> > sys     0m21.285s
> > after applying the patch
> > real    4m15.313s
> > user    3m42.598s
> > sys     0m18.485s
>
> --
> This message is automatically generated by JIRA.
> -
> If you think it was sent incorrectly contact one of the administrators:
> http://issues.apache.org/jira/secure/Administrators.jspa
> -
> For more information on JIRA, see: http://www.atlassian.com/software/jira
>
>
>


--
AJ Chen, PhD
http://web2express.org
Reply | Threaded
Open this post in threaded view
|

Re: [jira] Resolved: (NUTCH-395) Increase fetching speed

Sami Siren-2
from what version are you "upgrading" from? I guess pre rev. 464654?

If so, see [1] for additional info.

--
  Sami Siren

[1] http://wiki.apache.org/nutch/Upgrading_from_0%2e8%2ex_to_0%2e9

AJ Chen wrote:

> Sami,
> Thanks for resolving this serious issue.  I just updated my code from trunk
> and plan to test fetch speed. But ,there is a runtime error related to
> switching from UTF8 to Text. Since the error is from hadoop, how do I fix
> it?
>
> java.lang.ClassCastException: org.apache.hadoop.io.UTF8
>    at org.apache.nutch.crawl.Generato r$Selector.map(Generator.java:108)
>    at org.apache.hadoop.mapred.MapRunner.run(MapRunner.java:46)
>    at org.apache.hadoop.mapred.MapTask.run(MapTask.java:213)
>    at org.apache.hadoop.mapred.LocalJobRunner$Job.run(LocalJobRunner.java
> :105)
>
> Thanks,
> AJ
Reply | Threaded
Open this post in threaded view
|

Re: [jira] Commented: (NUTCH-395) Increase fetching speed

AJ Chen-2
In reply to this post by Nick Burch (Jira)
I checked out the code from trunk after Sami committed the change. I started
out a new crawl db and run several cycles of crawl sequentially on one linux
server. See below for the real numbers from my test.  The performance is
still poor because the crawler still spend too much time in reduce and
update operations.

#crawl cycle: topN=200000
2006-11-17 17:25:27,367 INFO  crawl.Generator - Generator: segment:
crawl/segments/20061117172527
2006-11-17 17:47:45,837 INFO  fetcher.Fetcher - Fetcher: segment:
crawl/segments/20061117172527
# 8 hours fetching ~200000 pages
2006-11-18 03:13:31,992 INFO  mapred.LocalJobRunner - 183644 pages, 5506
errors, 5.4 pages/s, 1043 kb/s,
# 4 hours doing "reduce"
2006-11-18 07:30:38,085 INFO  crawl.CrawlDb - CrawlDb update: starting
# 4 hours update db
2006-11-18 11:17:54,000 INFO  crawl.CrawlDb - CrawlDb update: done

#crawl sycle: topN=500,000 pages
2006-11-18 13:22:51,530 INFO  crawl.Generator - Generator: segment:
crawl/segments/20061118132251
2006-11-18 14:50:07,006 INFO  fetcher.Fetcher - Fetcher: segment:
crawl/segments/20061118132251
# fetching for 16 hours
2006-11-19 06:53:34,923 INFO  mapred.LocalJobRunner - 394343 pages, 19050
errors, 6.8 pages/s, 1439 kb/s,
# reduce for 11 hours
2006-11-19 17:49:15,778 INFO  crawl.CrawlDb - CrawlDb update: segment:
crawl/segments/20061118132251
# update db for 10 hours
2006-11-20 03:55:22,882 INFO  crawl.CrawlDb - CrawlDb update: done

#crawl cycle: topN=600,000 pages
2006-11-20 08:14:51,463 INFO  crawl.Generator - Generator: segment:
crawl/segments/20061120081451
2006-11-20 11:31:22,384 INFO  fetcher.Fetcher - Fetcher: segment:
crawl/segments/20061120081451
#fetching for 18 hours
2006-11-21 06:00:08,504 INFO  mapred.LocalJobRunner - 410078 pages, 26316
errors, 6.2 pages/s, 1257 kb/s,
#reduce for 11 hours
2006-11-21 17:26:38,213 INFO  crawl.CrawlDb - CrawlDb update: starting
#update for 13 hours
2006-11-22 06:25:48,592 INFO  crawl.CrawlDb - CrawlDb update: done


-AJ


On 11/13/06, Andrzej Bialecki (JIRA) <[hidden email]> wrote:

>
>     [
> http://issues.apache.org/jira/browse/NUTCH-395?page=comments#action_12449292]
>
> Andrzej Bialecki  commented on NUTCH-395:
> -----------------------------------------
>
> +1 - this patch looks good to me - if you could just fix the whitespace
> issues prior to committing, so that it conforms to the coding style ...
>
> > Increase fetching speed
> > -----------------------
> >
> >                 Key: NUTCH-395
> >                 URL: http://issues.apache.org/jira/browse/NUTCH-395
> >             Project: Nutch
> >          Issue Type: Improvement
> >          Components: fetcher
> >    Affects Versions: 0.9.0, 0.8.1
> >            Reporter: Sami Siren
> >         Assigned To: Sami Siren
> >         Attachments: nutch-0.8-performance.txt,
> NUTCH-395-trunk-metadata-only-2.patch, NUTCH-395-trunk-metadata-only.patch
> >
> >
> > There have been some discussion on nutch mailing lists about fetcher
> being slow, this patch tried to address that. the patch is just a quich hack
> and needs some cleaning up, it also currently applies to 0.8 branch and
> not trunk and it has also not been tested in large. What it changes?
> > Metadata - the original metadata uses spellchecking, new version does
> not (a decorator is provided that can do it and it should perhaps be used
> where http headers are handled but in most of the cases the functionality is
> not required)
> > Reading/writing various data structures - patch tries to do io more
> efficiently see the patch for details.
> > Initial benchmark:
> > A small benchmark was done to measure the performance of changes with a
> script that basically does the following:
> > -inject a list of urls into a fresh crawldb
> > -create fetchlist (10k urls pointing to local filesystem)
> > -fetch
> > -updatedb
> > original code from 0.8-branch:
> > real    10m51.907s
> > user    10m9.914s
> > sys     0m21.285s
> > after applying the patch
> > real    4m15.313s
> > user    3m42.598s
> > sys     0m18.485s
>
> --
> This message is automatically generated by JIRA.
> -
> If you think it was sent incorrectly contact one of the administrators:
> http://issues.apache.org/jira/secure/Administrators.jspa
> -
> For more information on JIRA, see: http://www.atlassian.com/software/jira
>
>
>


--
AJ Chen, PhD
Palo Alto, CA
http://web2express.org
Reply | Threaded
Open this post in threaded view
|

Re: [jira] Commented: (NUTCH-395) Increase fetching speed

Sami Siren-2
What kind of hardware are you running on? Your pages per sec ratio seems
very low to me.

How big was your crawldb when you started and how big was it at end?

What kind of filters and normalizers are you using?

--
  Sami Siren

AJ Chen wrote:

> I checked out the code from trunk after Sami committed the change. I
> started
> out a new crawl db and run several cycles of crawl sequentially on one
> linux
> server. See below for the real numbers from my test.  The performance is
> still poor because the crawler still spend too much time in reduce and
> update operations.
>
> #crawl cycle: topN=200000
> 2006-11-17 17:25:27,367 INFO  crawl.Generator - Generator: segment:
> crawl/segments/20061117172527
> 2006-11-17 17:47:45,837 INFO  fetcher.Fetcher - Fetcher: segment:
> crawl/segments/20061117172527
> # 8 hours fetching ~200000 pages
> 2006-11-18 03:13:31,992 INFO  mapred.LocalJobRunner - 183644 pages, 5506
> errors, 5.4 pages/s, 1043 kb/s,
> # 4 hours doing "reduce"
> 2006-11-18 07:30:38,085 INFO  crawl.CrawlDb - CrawlDb update: starting
> # 4 hours update db
> 2006-11-18 11:17:54,000 INFO  crawl.CrawlDb - CrawlDb update: done
>
> #crawl sycle: topN=500,000 pages
> 2006-11-18 13:22:51,530 INFO  crawl.Generator - Generator: segment:
> crawl/segments/20061118132251
> 2006-11-18 14:50:07,006 INFO  fetcher.Fetcher - Fetcher: segment:
> crawl/segments/20061118132251
> # fetching for 16 hours
> 2006-11-19 06:53:34,923 INFO  mapred.LocalJobRunner - 394343 pages, 19050
> errors, 6.8 pages/s, 1439 kb/s,
> # reduce for 11 hours
> 2006-11-19 17:49:15,778 INFO  crawl.CrawlDb - CrawlDb update: segment:
> crawl/segments/20061118132251
> # update db for 10 hours
> 2006-11-20 03:55:22,882 INFO  crawl.CrawlDb - CrawlDb update: done
>
> #crawl cycle: topN=600,000 pages
> 2006-11-20 08:14:51,463 INFO  crawl.Generator - Generator: segment:
> crawl/segments/20061120081451
> 2006-11-20 11:31:22,384 INFO  fetcher.Fetcher - Fetcher: segment:
> crawl/segments/20061120081451
> #fetching for 18 hours
> 2006-11-21 06:00:08,504 INFO  mapred.LocalJobRunner - 410078 pages, 26316
> errors, 6.2 pages/s, 1257 kb/s,
> #reduce for 11 hours
> 2006-11-21 17:26:38,213 INFO  crawl.CrawlDb - CrawlDb update: starting
> #update for 13 hours
> 2006-11-22 06:25:48,592 INFO  crawl.CrawlDb - CrawlDb update: done
>
>
> -AJ
>
>
> On 11/13/06, Andrzej Bialecki (JIRA) <[hidden email]> wrote:
>>
>>     [
>> http://issues.apache.org/jira/browse/NUTCH-395?page=comments#action_12449292]
>>
>>
>> Andrzej Bialecki  commented on NUTCH-395:
>> -----------------------------------------
>>
>> +1 - this patch looks good to me - if you could just fix the whitespace
>> issues prior to committing, so that it conforms to the coding style ...
>>
>> > Increase fetching speed
>> > -----------------------
>> >
>> >                 Key: NUTCH-395
>> >                 URL: http://issues.apache.org/jira/browse/NUTCH-395
>> >             Project: Nutch
>> >          Issue Type: Improvement
>> >          Components: fetcher
>> >    Affects Versions: 0.9.0, 0.8.1
>> >            Reporter: Sami Siren
>> >         Assigned To: Sami Siren
>> >         Attachments: nutch-0.8-performance.txt,
>> NUTCH-395-trunk-metadata-only-2.patch,
>> NUTCH-395-trunk-metadata-only.patch
>> >
>> >
>> > There have been some discussion on nutch mailing lists about fetcher
>> being slow, this patch tried to address that. the patch is just a
>> quich hack
>> and needs some cleaning up, it also currently applies to 0.8 branch and
>> not trunk and it has also not been tested in large. What it changes?
>> > Metadata - the original metadata uses spellchecking, new version does
>> not (a decorator is provided that can do it and it should perhaps be used
>> where http headers are handled but in most of the cases the
>> functionality is
>> not required)
>> > Reading/writing various data structures - patch tries to do io more
>> efficiently see the patch for details.
>> > Initial benchmark:
>> > A small benchmark was done to measure the performance of changes with a
>> script that basically does the following:
>> > -inject a list of urls into a fresh crawldb
>> > -create fetchlist (10k urls pointing to local filesystem)
>> > -fetch
>> > -updatedb
>> > original code from 0.8-branch:
>> > real    10m51.907s
>> > user    10m9.914s
>> > sys     0m21.285s
>> > after applying the patch
>> > real    4m15.313s
>> > user    3m42.598s
>> > sys     0m18.485s
>>
>> --
>> This message is automatically generated by JIRA.
>> -
>> If you think it was sent incorrectly contact one of the administrators:
>> http://issues.apache.org/jira/secure/Administrators.jspa
>> -
>> For more information on JIRA, see: http://www.atlassian.com/software/jira
>>
>>
>>
>
>

Reply | Threaded
Open this post in threaded view
|

Re: [jira] Commented: (NUTCH-395) Increase fetching speed

AJ Chen-2
Linux box, opteron 2Ghz, 2GB RAM, DSL download bandwidth up to 5mbps.

This is a new crawldb, crawling on 4000 selected sites, total ~1 million
pages fetched after last run.

use default regex-urlfilter.txt except for :
-(?i)\.(ai|asf|au|avi|bz2|bin|bmp|c|class|css|dmg|doc|dot|dvi|eps|exe|gif|gz|h|hqx|ico|iso|jar|java|jnlp|jpeg|jpg|lha|md5|mov|
mp3|mp4|mpg|msi|ogg|png|pps|ppt|ps|psd|ram|ris|rm|rpm|rss|rtf|sit|swf|tar|tbz|tbz2|tgz|tif|wav|wmf|wmv|xls|z|zip)\)?$
-[*!@#]

additional filter to limit urls to the selected domains  (hashtable
implementation)

plugins:
protocol-http|urlfilter-regex|parse-(text|html)|index-basic|query-(basic|site|url)|summary-basic|scoring-opic

use default org.apache.nutch.net.URLNormalizer

thanks for helping,
AJ


parse only html and text

On 11/22/06, Sami Siren <[hidden email]> wrote:

>
> What kind of hardware are you running on? Your pages per sec ratio seems
> very low to me.
>
> How big was your crawldb when you started and how big was it at end?
>
> What kind of filters and normalizers are you using?
>
> --
>   Sami Siren
>
> AJ Chen wrote:
> > I checked out the code from trunk after Sami committed the change. I
> > started
> > out a new crawl db and run several cycles of crawl sequentially on one
> > linux
> > server. See below for the real numbers from my test.  The performance is
> > still poor because the crawler still spend too much time in reduce and
> > update operations.
> >
> > #crawl cycle: topN=200000
> > 2006-11-17 17:25:27,367 INFO  crawl.Generator - Generator: segment:
> > crawl/segments/20061117172527
> > 2006-11-17 17:47:45,837 INFO  fetcher.Fetcher - Fetcher: segment:
> > crawl/segments/20061117172527
> > # 8 hours fetching ~200000 pages
> > 2006-11-18 03:13:31,992 INFO  mapred.LocalJobRunner - 183644 pages, 5506
> > errors, 5.4 pages/s, 1043 kb/s,
> > # 4 hours doing "reduce"
> > 2006-11-18 07:30:38,085 INFO  crawl.CrawlDb - CrawlDb update: starting
> > # 4 hours update db
> > 2006-11-18 11:17:54,000 INFO  crawl.CrawlDb - CrawlDb update: done
> >
> > #crawl sycle: topN=500,000 pages
> > 2006-11-18 13:22:51,530 INFO  crawl.Generator - Generator: segment:
> > crawl/segments/20061118132251
> > 2006-11-18 14:50:07,006 INFO  fetcher.Fetcher - Fetcher: segment:
> > crawl/segments/20061118132251
> > # fetching for 16 hours
> > 2006-11-19 06:53:34,923 INFO  mapred.LocalJobRunner - 394343 pages,
> 19050
> > errors, 6.8 pages/s, 1439 kb/s,
> > # reduce for 11 hours
> > 2006-11-19 17:49:15,778 INFO  crawl.CrawlDb - CrawlDb update: segment:
> > crawl/segments/20061118132251
> > # update db for 10 hours
> > 2006-11-20 03:55:22,882 INFO  crawl.CrawlDb - CrawlDb update: done
> >
> > #crawl cycle: topN=600,000 pages
> > 2006-11-20 08:14:51,463 INFO  crawl.Generator - Generator: segment:
> > crawl/segments/20061120081451
> > 2006-11-20 11:31:22,384 INFO  fetcher.Fetcher - Fetcher: segment:
> > crawl/segments/20061120081451
> > #fetching for 18 hours
> > 2006-11-21 06:00:08,504 INFO  mapred.LocalJobRunner - 410078 pages,
> 26316
> > errors, 6.2 pages/s, 1257 kb/s,
> > #reduce for 11 hours
> > 2006-11-21 17:26:38,213 INFO  crawl.CrawlDb - CrawlDb update: starting
> > #update for 13 hours
> > 2006-11-22 06:25:48,592 INFO  crawl.CrawlDb - CrawlDb update: done
> >
> >
> > -AJ
> >
> >
> > On 11/13/06, Andrzej Bialecki (JIRA) <[hidden email]> wrote:
> >>
> >>     [
> >>
> http://issues.apache.org/jira/browse/NUTCH-395?page=comments#action_12449292
> ]
> >>
> >>
> >> Andrzej Bialecki  commented on NUTCH-395:
> >> -----------------------------------------
> >>
> >> +1 - this patch looks good to me - if you could just fix the whitespace
> >> issues prior to committing, so that it conforms to the coding style ...
> >>
> >> > Increase fetching speed
> >> > -----------------------
> >> >
> >> >                 Key: NUTCH-395
> >> >                 URL: http://issues.apache.org/jira/browse/NUTCH-395
> >> >             Project: Nutch
> >> >          Issue Type: Improvement
> >> >          Components: fetcher
> >> >    Affects Versions: 0.9.0, 0.8.1
> >> >            Reporter: Sami Siren
> >> >         Assigned To: Sami Siren
> >> >         Attachments: nutch-0.8-performance.txt,
> >> NUTCH-395-trunk-metadata-only-2.patch,
> >> NUTCH-395-trunk-metadata-only.patch
> >> >
> >> >
> >> > There have been some discussion on nutch mailing lists about fetcher
> >> being slow, this patch tried to address that. the patch is just a
> >> quich hack
> >> and needs some cleaning up, it also currently applies to 0.8 branch and
> >> not trunk and it has also not been tested in large. What it changes?
> >> > Metadata - the original metadata uses spellchecking, new version does
> >> not (a decorator is provided that can do it and it should perhaps be
> used
> >> where http headers are handled but in most of the cases the
> >> functionality is
> >> not required)
> >> > Reading/writing various data structures - patch tries to do io more
> >> efficiently see the patch for details.
> >> > Initial benchmark:
> >> > A small benchmark was done to measure the performance of changes with
> a
> >> script that basically does the following:
> >> > -inject a list of urls into a fresh crawldb
> >> > -create fetchlist (10k urls pointing to local filesystem)
> >> > -fetch
> >> > -updatedb
> >> > original code from 0.8-branch:
> >> > real    10m51.907s
> >> > user    10m9.914s
> >> > sys     0m21.285s
> >> > after applying the patch
> >> > real    4m15.313s
> >> > user    3m42.598s
> >> > sys     0m18.485s
> >>
> >> --
> >> This message is automatically generated by JIRA.
> >> -
> >> If you think it was sent incorrectly contact one of the administrators:
> >> http://issues.apache.org/jira/secure/Administrators.jspa
> >> -
> >> For more information on JIRA, see:
> http://www.atlassian.com/software/jira
> >>
> >>
> >>
> >
> >
>
>


--
AJ Chen, PhD
Palo Alto, CA
http://web2express.org