grady@scam.Berkeley.EDU (Steven Grady) (02/22/90)
(I've seen thiis problem mentioned before. I never saw an answer.) Weep for me, world. Two thirds of my nn data files got zeroed somehow. This makes it very very hard to deal with nn. I don't understand why this is happening -- I just switched sites, and it never occurred at the old site. The only way I know to fix it is to re-initialize the database, which (since it requests articles via NNTP) takes at least 7 or 8 hours, and even then there is no guarantee that same problem will not occur. When I originally installed nn at this site, after the initialization, many of the files were zeroed. The only way to get the database up to speed was to leave the master running for a week, allowing it to incrementally fill in the database. I recently restarted the master, so I guess that's why the files got zeroed. Also, the db directory is not on the same machine as master -- they are nfs-linked. (This was not the case at my old site.) Reluctantly, I am switching back to rn for the moment. I'd prefer to use nn, so can someone please tell me how to fix this travesty? Steven grady@postgres.berkeley.edu ...!ucbvax!grady
mr@cica.cica.indiana.edu (Michael Regoli) (02/23/90)
In article <22292@pasteur.Berkeley.EDU> grady@scam.Berkeley.EDU (Steven Grady) writes: >(I've seen thiis problem mentioned before. I never saw an answer.) >Weep for me, world. Two thirds of my nn data files got zeroed >somehow. This makes it very very hard to deal with nn. Well, this isn't unique then. We just had a helluva time cleaning zero length files out of our spool as well. Apparently they were coming that way. (This, on top of an nn upgrade to patchlevel 10...) I'm glad I'm not the only one going crazy. However, I'm happy to report that all's well here in the Hinterland. -- michael regoli mr@cica.indiana.edu regoli@iubacs.bitnet ...rutgers!iuvax!cica!mr things fall apart--it's scientific.