Hi Jeff et al, As you suggested I compiled up the latest stable 2.6.39.1 kernel from kernel.org. I copied the .config file from 2.6.32-5-amd64 stock debian kernel I was running before and didn't make any changes. I've been running the new kernel about a week now and the error messages are gradually increasing. Not to the point they were with the old kernel as I typically get one every few minutes. Below is a sample of messages from /var/log/kern.log this morning. Jun 20 04:28:55 fujimoto kernel: [414833.777315] CIFS VFS: Send error in FindClose = -9 Jun 20 04:48:25 fujimoto kernel: [416001.742074] CIFS VFS: Unexpected lookup error -112 Jun 20 07:41:16 fujimoto kernel: [426354.188394] CIFS VFS: Send error in read = -11 Jun 20 07:41:45 fujimoto kernel: [426384.120676] CIFS VFS: Send error in read = -11 Jun 20 07:41:45 fujimoto kernel: [426384.120700] CIFS VFS: Send error in read = -11 Jun 20 07:43:15 fujimoto kernel: [426473.247188] CIFS VFS: Send error in read = -11 Jun 20 07:45:45 fujimoto kernel: [426622.971670] CIFS VFS: Send error in Close = -11 Jun 20 07:46:15 fujimoto kernel: [426652.919689] CIFS VFS: Send error in read = -11 Jun 20 07:47:15 fujimoto kernel: [426712.818888] CIFS VFS: Send error in read = -11 Jun 20 07:47:45 fujimoto kernel: [426742.763649] CIFS VFS: Send error in read = -11 Jun 20 07:48:45 fujimoto kernel: [426802.660220] CIFS VFS: Send error in read = -11 Jun 20 07:50:14 fujimoto kernel: [426892.176104] CIFS VFS: Send error in read = -11 Jun 20 07:50:14 fujimoto kernel: [426892.176166] CIFS VFS: Send error in read = -11 Jun 20 07:52:43 fujimoto kernel: [427040.913850] CIFS VFS: Unexpected SMB signature Jun 20 07:53:13 fujimoto kernel: [427070.679066] CIFS VFS: Send error in read = -11 Jun 20 07:54:13 fujimoto kernel: [427130.156587] CIFS VFS: Send error in read = -11 Jun 20 07:55:12 fujimoto kernel: [427189.226655] CIFS VFS: Send error in read = -11 I'll recompile the kernel with a different config if someone has some suggestions. Shane McColman PEng. Quality Assurance Engineer Dynamic Source Manufacturing Inc. Calgary AB, Canada -----Original Message----- From: Jeff Layton [mailto:jlayton@xxxxxxxxxxxxxxx] On Behalf Of Jeff Layton Sent: Wednesday, June 08, 2011 2:44 PM To: Shane McColman Cc: linux-cifs@xxxxxxxxxxxxxxx Subject: Re: CIFS VFS errors On Wed, 8 Jun 2011 13:56:10 -0600 "Shane McColman" <smccolman@xxxxxxxxxxxxxxxxxxxx> wrote: > Hi Jeff, > > > > I had tried to send out the below message last week. It never bounced. Not > sure if anyone got it. I could really use some help. > > Nope, never saw it. Maybe kernel.org spam filtering ate it? > > > _____ > > From: Shane McColman [mailto:smccolman@xxxxxxxxxxxxxxxxxxxx] > Sent: Thursday, June 02, 2011 9:06 AM > To: 'linux-cifs@xxxxxxxxxxxxxxx' > Subject: CIFS VFS errors > > > > First some background: > > > > I'm using CIFS on Debian 6.0.1 with kernel 2.6.32-5-amd64 > > > > I've mounted a shared folder from a windows XP home machine that contains > over 4000 files and over 2000 directories. This machine is the file > repository for a group of equipment we have on our factory floor. This > system is supplied by the manufacturer of the equipment, so it cannot be > easily changed. I could back up the files and folders to reduce the size of > the directory, but outside of that I'm stuck with it as is. > > > > I've written some code that monitors the directory and looks for new log > files being created, whenever a new file is created it starts up a new > thread that monitors the file for modifications and then parses the > modifications into a database. At the end of each month these log files > will be not be modified anymore so after the clock rolls over on the first > of the month all the active threads are shutdown and new ones started as new > files are added. Usually by the end of month there are about 300 threads > running. > > > > The problem: > > > > By about mid month I starting getting errors every 15 sec that look like > this (/var/log/kern.log) > > > > May 31 23:59:40 fujimoto kernel: [921847.411084] CIFS VFS: No response for > cmd 50 mid 30108 > > May 31 23:59:40 fujimoto kernel: [921847.411183] CIFS VFS: No response for > cmd 50 mid 30317 > > May 31 23:59:40 fujimoto kernel: [921847.411188] CIFS VFS: No response for > cmd 50 mid 30265 > > May 31 23:59:40 fujimoto kernel: [921847.411191] CIFS VFS: Send error in > SessSetup = -11 > > May 31 23:59:40 fujimoto kernel: [921847.411343] CIFS VFS: No response for > cmd 50 mid 30313 > > May 31 23:59:40 fujimoto kernel: [921847.411834] CIFS VFS: Send error in > SessSetup = -11 > > May 31 23:59:40 fujimoto kernel: [921847.411893] CIFS VFS: Send error in > SessSetup = -11 > > May 31 23:59:40 fujimoto kernel: [921847.411951] CIFS VFS: Send error in > SessSetup = -11 > > May 31 23:59:40 fujimoto kernel: [921847.412003] CIFS VFS: Send error in > SessSetup = -11 > > May 31 23:59:40 fujimoto kernel: [921847.415091] CIFS VFS: No response for > cmd 50 mid 30119 > > May 31 23:59:40 fujimoto kernel: [921847.415115] CIFS VFS: No response for > cmd 50 mid 30250 > > > > There are a lot more (log file grows by 25M a day) but this is a good > sample. I first noticed this because I was getting file IO errors with my > script and since then have rewritten my code to copy the file to the server > before parsing and added error handling into my code so that it will try 10 > times if unsuccessful. By month end it was taking 3 to 4 attempts to copy > files that were each under 250K. > > > > As we rolled into June and the threads were shut down the errors are much > less > > > > Jun 1 04:15:29 fujimoto kernel: [937168.611757] CIFS VFS: No response for > cmd 114 mid 7014 > > Jun 1 04:15:30 fujimoto kernel: [937169.756306] CIFS VFS: No response for > cmd 114 mid 7015 > > Jun 1 04:15:31 fujimoto kernel: [937170.763869] CIFS VFS: No response for > cmd 114 mid 7016 > > Jun 1 04:15:36 fujimoto kernel: [937175.822892] CIFS VFS: Unexpected > lookup error -112 > > Jun 1 04:15:37 fujimoto kernel: [937177.028741] CIFS VFS: No response for > cmd 114 mid 7022 > > Jun 1 04:15:38 fujimoto kernel: [937178.038878] CIFS VFS: No response for > cmd 114 mid 7023 > > Jun 1 04:15:38 fujimoto kernel: [937178.038886] CIFS VFS: Unexpected > lookup error -112 > > Jun 1 04:15:53 fujimoto kernel: [937193.012420] CIFS VFS: No response for > cmd 114 mid 7024 > > Jun 1 04:15:53 fujimoto kernel: [937193.012436] CIFS VFS: Unexpected > lookup error -112 > > Jun 1 04:16:03 fujimoto kernel: [937202.994543] CIFS VFS: Unexpected > lookup error -112 > > Jun 1 04:16:13 fujimoto kernel: [937212.976749] CIFS VFS: Unexpected > lookup error -112 > > Jun 1 04:50:12 fujimoto kernel: [939247.838590] CIFS VFS: No response for > cmd 50 mid 50347 > > Jun 1 04:50:13 fujimoto kernel: [939248.988539] CIFS VFS: No response for > cmd 50 mid 50348 > > Jun 1 10:54:00 fujimoto kernel: [961037.316672] CIFS VFS: No response for > cmd 114 mid 19781 > > Jun 1 10:54:01 fujimoto kernel: [961038.326872] CIFS VFS: No response for > cmd 114 mid 19782 > > Jun 1 10:54:02 fujimoto kernel: [961039.337069] CIFS VFS: No response for > cmd 114 mid 19783 > > > > My first question is "Is this a configuration error?" and if so can someone > give me some advice. > > No, probably not a config error. Those errors basically indicate that the server just wasn't responding in a timely fashion. There has been a lot of work in this area over the last couple of kernel releases -- we only time out calls now when the server stops responding to SMB echo requests. You may want to try something 2.6.39-ish. -- Jeff Layton <jlayton@xxxxxxxxx> -- To unsubscribe from this list: send the line "unsubscribe linux-cifs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
