RE: [squid-users] squid-2.5.stable9 aufs issue

From: Elsen Marc <[email protected]>
Date: Fri, 1 Apr 2005 13:44:50 +0200

 
>
> Hello,
>
> we are working with a production squid-2.5.stable9 on a Red Hat
> Enterprise Linux AS release 4 (Nahant), redhat prepared
> 2.6.9-5 kernel,
> and are getting the following messages, aufs related, on cache.log..
>
> 2005/04/01 11:24:36| diskHandleWrite: FD 216: disk write error: (30)
> Read-only file system
> 2005/04/01 11:24:36| storeAufsWriteDone: got failure (-1)
> 2005/04/01 11:24:36| storeSwapOutFileClosed: dirno 0,
> swapfile 000158F5,
> errflag=-1
> (30) Read-only file system
> 2005/04/01 11:24:36| storeAufsOpenDone: (30) Read-only file system
> 2005/04/01 11:24:36| /cache/01/58/000158F8
> 2005/04/01 11:24:36| storeSwapOutFileClosed: dirno 0,
> swapfile 000158F8,
> errflag=-1
> (30) Read-only file system
> 2005/04/01 11:24:36| storeAufsOpenDone: (30) Read-only file system
> 2005/04/01 11:24:36| /cache/01/58/000158FA
> 2005/04/01 11:24:36| storeSwapOutFileClosed: dirno 0,
> swapfile 000158FA,
> errflag=-1
> (30) Read-only file system
> 2005/04/01 11:24:36| ctx: enter level 0:
> 'http://www.cadenaser.com/player/acontinuacion_radio.html'
> 2005/04/01 11:24:36| diskHandleWrite: FD 8: disk write error: (30)
> Read-only file system
> 2005/04/01 11:24:36| storeDirWriteCleanLogs: Starting...
> 2005/04/01 11:24:36| storeDirWriteCleanStart:
> /cache/swap.state.clean:
> open: (30) Read-only file system
> 2005/04/01 11:24:36| storeDirWriteCleanStart: Current swap
> logfile not
> replaced.
> 2005/04/01 11:24:36| log.clean.start() failed for dir #0
> 2005/04/01 11:24:36| WARNING: Closing open FD 12
> 2005/04/01 11:24:36| Finished. Wrote 0 entries.
> 2005/04/01 11:24:36| Took 0.0 seconds ( 0.0 entries/sec).
> FATAL: Write failure -- check your disk space and cache.log
> Squid Cache (Version 2.5.STABLE9): Terminated abnormally.
>
> Configuration for squid is the following..
>
> ./configure \
> --prefix=/usr/local/squid-2.5.STABLE9 \
> --disable-wccp \
> --enable-ssl \
> --with-openssl=/usr/local/ssl \
> --enable-cache-digests \
> --enable-default-err-language=Spanish \
> --enable-err-languages="Spanish English" \
> --disable-ident-lookups \
> --disable-hostname-checks \
> --enable-underscores \
> --enable-async-io
>
> We had prepared another pool /cache2 for cache writing with ufs, and
> moved squid.conf configuration from
>
> cache_dir aufs /cache 20480 16 256
> #cache_dir ufs /cache2 20480 16 256
>
> #cache_dir aufs /cache 20480 16 256
> cache_dir ufs /cache2 20480 16 256
>
> Everything is working well now.. Any idea for the error
> above? Should I
> report a bug o something?..
>
 
 Seems like the original partition used was mounted read-only.
 Won't do any good for squid.
 Make sure that files can be created in the intended cache dir(s).
 Verify this manually by making some files in the intended cache dir(s),
 do this as the user squid is intended to run under.

 M.
Received on Fri Apr 01 2005 - 04:46:04 MST

This archive was generated by hypermail pre-2.1.9 : Sun May 01 2005 - 12:00:03 MDT