Subj : fEUTIL LINK
To   : Charles Cruden
From : mark lewis
Date : Tue Sep 17 2002 09:54 am

>> I like to keep a lot of messages in some areas...
>> but looks like "feutil LINK" is limited?
>> fn_sysop now contains 21574 messages -> too much
>> nodes.024 contains 14877 messages -> too much
>> enet.sysop contains 8017 messages -> too much
>> skypoints contains 4996 messages -> this worked still :-)
>>
>> are t here other JAM linkers avaible that don't have this problem?

 CC> I've seen the same sort of problem here, only with FAR fewer
 CC> messages.  Usually once a message base hits 1000-1500
 CC> messages, messages start to disappear.  If I leave things too
 CC> long, the index file gets corrupted and pointers start to go
 CC> into the middle of messages instead of the start.

yuck... once a pointer gets out of whack, there are (generally) at least two 
messages tagged out... it is possible to repair but definitely takes a tool 
and/or some serious hexediting and study... i had something way back that 
is/was extremely crude that allowed me to repair some things but it was manual 
and could lead to further corruption...

 CC> (This is with message bases where messages are not allowed to
 CC> expire for any reason.  The drive(s) involved have plenty of
 CC> free space: 400M to 600M.)

that is wierd... i had that problem a few times back during fd, fe, ra beta 
testing times but not since then... ra does, even in the latest beta, (still) 
have a message delete bug if users are allowed to delete messages... but i 
haven't seen this type of error in a long long long time... i've several bases 
here with huge numbers of messages in them...

i wonder... are you using a filesystem that (automatically) defrags itself or 
are you using one of the FAT based formats? my system is OS/2 and runs with all 
HPFS formatted drives... HPFS doesn't break files into pieces and scatter them 
all over the place like FAT does...

just something to think about...

)\/(ark

 
* Origin: (1:3634/12)

.