Tuesday, July 24, 2007

users feedback on sharehound 1.2.3alpha

Feel free to leave a comment on your experience with this release, sharehound in general or whatever :) Any language is OK. Feature requests are welcome.

If you wish to post anonymously please leave some nick so that I could comment back (as blogger.com doesn't support hierarchical comments).

11 comments:

Shadow said...

Windows 2000 server+JDK 1.6.0_02+Tomcat 5.5.23+.NET Framework 2.0 (for msvcr71.dll)ShareHound 1.2.3 Alpha=it works :)

Shadow said...

How can I change schedule in quartz-config.xml to run jobs every night at particular time ?

Shadow said...

And another problem. It was in the previous release too. If I had changed "default.page.size=15" in "search.properties" file to a greater value (50 for example) then I got 50 results on a page but only 15 PATH-names :-(

Artem Vasiliev said...

Cool!

.net framework, that seem to be quite an overhead. (I think) I don't have it at my corporate Sharehound instance btw. I use jdk1.6.0_02 (it has msvcr71.dll, strrange that yours doesn't) and tomcat-6.0.10 here.

Shadow said...

Windows 2000 server+JDK 1.6.0_02+Tomcat 5.5.23+.NET Framework 2.0 (for msvcr71.dll)ShareHound 1.2.3 Alpha=it works :)

Artem Vasiliev said...

Well, good question, I guess I should include an example in sh, or make it default. Quartz supports cron-expressions, exact syntax is described here.

For example, this trigger will fire IndexLocalSubnet job at 01:00 -

<trigger>
<cron>
<name>IndexLocalSubnetTrigger</name>
<group>DEFAULT</group>
<job-name>IndexLocalSubnet</job-name>
<job-group>DEFAULT</job-group>
<cron-expression>0 0 1 ? * *</cron-expression>
</cron>
</trigger>

Shadow said...

How can I change schedule in quartz-config.xml to run jobs every night at particular time ?

Artem Vasiliev said...

Sorry I don't understand. Where are those PATH-names in sharehound?

Shadow said...

...but only 15 PATH-names :-(

Artem Vasiliev said...

Wow, just saw it myself, really disgusting! thanks :)

Shadow said...

...but only 15 PATH-names :-(

Anonymous said...

Ambrosio here:

Hi, I was able to get 1.2.3alpha running on the latest Fedora 9 respin. Seems to be working well for
FTP repositories. How do I specify the indexer/crawler to search only specific FTP folders in a single IP?

Artem Vasiliev said...

Hi Ambrosio!

Not sure if it will work, but you can try to specify host with path in SingleRootCrawlJob at quartz-config.xml - take example from .quartz-config-examples/quartz-config.xml.single.ftp.

E.g. ftp://myhost.com/pub/distros

Anonymous said...

deployment requirements should state java 1.6 because of java.lang.String.isEmpty..
but after i upgraded java,
specified huge enough XX:MaxPermSize,
replaced the used lucene-core jar file with version 2.4 (it handles restarts a lot better),
and changed the logging to a little less verbose level...
it looks nice, thx..
(ok, i haven't had a tomcat restart that didn't end in killing since, but considering all the threads, it's to be expected i guess)
do wish there weren't multiple entries for laptops that get assigned different ip-s though.

tomcat 5.5.9 on debian 4.0 64-bit

x

Artem Vasiliev said...

Thanks x! useful hints.

>do wish there weren't multiple entries for laptops that get assigned different ip-s though.

yeah.. the problem is I don't get this in environment I use Sh :) So I just don't get to this. Actually I don't get to any Sh support lately :(

I wish it was made in Ruby, just can't think of improving all this Java code now :)

May be using host IP or name should become an attribute of scan job - in different situations we could prefer this or that. Or maybe, we should just use name as key - people don't tend to change their computer names often.