> At 11:42 AM -0800 11/16/01, Bob Okumura wrote:
> >Enter value for words: lake
> >WordDB: open : Cannot allocate memory
> >WordDB: environment not yet opened
> 
> Well, I'd ask how much memory you have available and how big your 
> databases are. Do you expect that a query for "lake" will return a 
> lot of results?
> -- 
> --
> -Geoff Hutchison

Hi Geoff, sorry for the delay in responding (out of town, 
etc.) -- and very sorry about not being clearer... 

The database I was using was a test database with only 
5 or so pages indexed, and with a number of "lake" words 
in them (as verified by htdump).  All the htdig db files 
were well under 100k each.

I've made some progress in narrowing down the problem, 
but have hit another roadblock.

After some trial-and-error, I executed the "unlimit" 
command at the command line and got ./htsearch to 
return several valid matching results and without returning 
the "cannot allocate memory" message.

HOWEVER, when I now try to execute htsearch via the 
web browser form (and therefore through the Apache web 
server), it results in a "No matches were found for 'lake'." 
message similar to my original problem at the command 
line.  Also, the following error messages showed up in 
my error_log file:

WordDB: open : Cannot allocate memory
WordDB: environment not yet opened
WordDB: DB->cursor: method meaningless before open
WordDB: DB->cursor: method meaningless before open
WordDB: DB->cursor: method meaningless before open
WordDB: DB->cursor: method meaningless before open

Checking the resources available through the web server 
gateway reveals the following: 

coredumpsize    unlimited
cputime         unlimited
datasize        16384 kbytes
filesize        unlimited
maxproc         96
memorylocked    85425 kbytes
memoryuse       256272 kbytes
openfiles       128
stacksize       2048 kbytes

In contrast, the resources available after issuing the 
"unlimit" command at the command line (and which 
resulted in a successful htsearch) were:

coredumpsize    unlimited
cputime         unlimited
datasize        65536 kbytes
filesize        unlimited
maxproc         96
memorylocked    256276 kbytes
memoryuse       256272 kbytes
openfiles       13196
stacksize       65536 kbytes

Am I out of luck trying to get htsearch to run within the 
memory, data size, and stack size limits of the web 
server?  Any suggestions on what I might try to find a 
workaround?  Being on a virtual server, the server 
resources are not within my control.

Thanks in advance,

Bob Okumura




_______________________________________________
htdig-general mailing list <[EMAIL PROTECTED]>
To unsubscribe, send a message to <[EMAIL PROTECTED]> with a 
subject of unsubscribe
FAQ: http://htdig.sourceforge.net/FAQ.html

Reply via email to