Hi guys, This just happened to me again, on snv_112. In case it helps, this is the entire content of /var/adm/messages:
# cat /var/adm/messages Apr 29 03:57:44 rob-039v idmap[1965]: [ID 696364 daemon.error] Degraded operation (Global Catalog servers not configured/discoverable). If you are running an SMB server in workgroup mode, or if you're not running an SMB server, then you can ignore this message Apr 29 04:24:06 rob-039v idmap[1965]: [ID 643943 daemon.notice] Normal operation restored Apr 29 07:48:59 rob-039v smbsrv: [ID 138215 kern.notice] NOTICE: smbd[ROBINSONS\Ross Smith]: admi share not found Apr 29 07:49:17 rob-039v last message repeated 1 time Apr 29 09:52:54 rob-039v smbsrv: [ID 138215 kern.notice] NOTICE: smbd[ROBINSONS\Ross Smith]: admi share not found Apr 29 10:10:29 rob-039v smbsrv: [ID 138215 kern.notice] NOTICE: smbd[ROBINSONS\Ross Smith]: admi share not found Apr 29 10:10:39 rob-039v last message repeated 1 time Apr 29 10:25:57 rob-039v smbsrv: [ID 138215 kern.notice] NOTICE: smbd[NT Authority\Anonymous]: admin access denied: IPC only Apr 29 10:25:57 rob-039v last message repeated 1 time Apr 29 10:29:48 rob-039v smbsrv: [ID 138215 kern.notice] NOTICE: smbd[NT Authority\Anonymous]: admin access denied: IPC only Apr 29 10:29:49 rob-039v last message repeated 1 time Apr 29 10:35:19 rob-039v smbsrv: [ID 138215 kern.notice] NOTICE: smbd[NT Authority\Anonymous]: admin access denied: IPC only Apr 29 10:41:07 rob-039v last message repeated 7 times Apr 29 10:41:54 rob-039v smbsrv: [ID 138215 kern.notice] NOTICE: smbd[NT Authority\Anonymous]: admin access denied: IPC only Apr 29 10:42:08 rob-039v last message repeated 3 times Apr 29 11:27:03 rob-039v smbsrv: [ID 138215 kern.notice] NOTICE: smbd[NT Authority\Anonymous]: admin access denied: IPC only However, this time I can close this windows explorer window and open a new one with the same symptoms. There are no new messages in the log though. It's still restricted to one computer though, connecting from another machine works fine. Is there anything I can look for in the logs that might help you track down the cause of this? Ross On Wed, Feb 4, 2009 at 8:45 PM, Alan M Wright <a...@sun.com> wrote: > Filed CR 6801332 Intermittent file change notification > > Alan > > On 02/04/09 08:58, Ross Smith wrote: >> >> Hi guys, >> >> Well, it's been a while but I've just had another example of this happen. >> >> This time it was on a snv_106 build, and the time is synchronised >> perfectly with the domain (I have ntpdate running as a cron job every >> 5 minutes, and checked it manually when this happened). >> >> I had just moved a whole bunch of files using cut & paste in windows >> explorer. When the move operation finished, the original folder was >> still visible until I hit F5 to manually refresh the display. >> >> I had two explorer windows open at the time. The new files had all >> appeared in the second window, but the main folder I was moving had >> not been removed from the other window. >> >> I've never seen explorer need refreshing from a windows server, and it >> even works fine in Solaris the majority of the time, but there's >> definitely an intermittent bug of some kind affecting the file change >> notifications. >> >> In fact, now I'm writing this, if I leave these two folders open, I'm >> getting problems with file change notification in one, but the other >> works perfectly. >> >> In the window with problems, if I right-click to create a new folder, >> I get exactly the same effect as before - nothing appears. After I >> hit F5, I can see the folder and rename it, but the rename doesn't >> show up either until I hit F5 again. >> >> However, in the second window, creating new folders works perfectly. >> >> I get the same effect moving files. They appear automatically in one >> window, but don't disappear from the other. >> >> If I close the first window down, and open a new explorer window to >> the same location, it works perfectly in both windows. >> >> So whatever this is, it's intermittent in how it starts, but >> persistent once the problem has occurred, and it's specific to an >> individual instance of windows explorer. >> >> Ross >> >> >> >> On Tue, Dec 9, 2008 at 6:42 AM, Brent Jones <br...@servuhome.net> wrote: >>> >>> On Thu, Dec 4, 2008 at 6:30 AM, Ross Smith <myxi...@googlemail.com> >>> wrote: >>>> >>>> Well, it's repeatable, but I've no idea how lol. >>>> >>>> Just had exactly the same happen today after doing a lot of testing on >>>> the server, but despite spending another hour on it, I can't reproduce >>>> the problem. >>>> >>>> These machines are suffering from the VMware clock skew issue, so it >>>> might be related to that. Probably safe to ignore this for now and >>>> I'll raise this again if I ever find what's actually causing it. >>>> >>>> Ross >>>> >>>> >>>> >>>> On Wed, Dec 3, 2008 at 7:38 PM, Ross Smith <myxi...@googlemail.com> >>>> wrote: >>>>> >>>>> Hey guys, >>>>> >>>>> It looks like you're right, it's working perfectly every time now. >>>>> >>>>> I did the original testing on my own XP SP3 workstation, but I've just >>>>> gone and repeated the test on a whole bunch of machines (SP2 and SP3), >>>>> and it worked fine on all of them, and also now works fine on the >>>>> original workstation. >>>>> >>>>> I have been doing a lot of testing on this pool (disconnecting iscsi >>>>> drives and working in degraded mode to test recovery), which has been >>>>> causing a few CIFS errors as things timeout. I'll see if I can get it >>>>> to reproduce this problem. >>>>> >>>>> Ross >>>>> >>>>> >>>>>> On Wed, Dec 3, 2008 at 7:04 PM, Afshin Salek <afshin.ardak...@sun.com> >>>>>> wrote: >>>>>>> >>>>>>> Ross, >>>>>>> >>>>>>> What Windows OS are you using? >>>>>>> Have you tried different Windows systems to see if the problem still >>>>>>> occurs? >>>>>>> >>>>>>> Afshin >>>>>>> >>>>>>> Alan M Wright wrote: >>>>>>>> >>>>>>>> On 12/03/08 10:42, Alan M Wright wrote: >>>>>>>>> >>>>>>>>> On 12/03/08 01:43, Ross wrote: >>>>>>>>>> >>>>>>>>>> I'm not sure what's causing this, but I'm finding that the Solaris >>>>>>>>>> CIFS >>>>>>>>>> shares don't update automatically like windows shares do. >>>>>>>>>> >>>>>>>>>> A couple of examples: >>>>>>>>>> >>>>>>>>>> If I right-click the folder and create a new file: >>>>>>>>>> - In windows, the new file appears, is selected, and I can simply >>>>>>>>>> type >>>>>>>>>> the name. When I press enter it's renamed and appears in the >>>>>>>>>> window. >>>>>>>>>> - On a solaris share, no file appears until I manually refresh. I >>>>>>>>>> then >>>>>>>>>> have to select the file manually and rename it. However, after >>>>>>>>>> renaming it >>>>>>>>>> the file still appears with the old name until I hit refresh >>>>>>>>>> again. If I >>>>>>>>>> try to delete the renamed file before hitting refresh I get an >>>>>>>>>> error "Cannot >>>>>>>>>> delete file: Cannot read from source file or disk." >>>>>>>>>> >>>>>>>>>> Similarly, if I drag a file into a Solaris share, the file does >>>>>>>>>> not >>>>>>>>>> appear until I hit F5 to refresh the window. >>>>>>>>>> >>>>>>>>>> This will be a particular problem if we roll this out as Windows >>>>>>>>>> doesn't >>>>>>>>>> have a refresh button in its explorer windows, and our users will >>>>>>>>>> almost >>>>>>>>>> certainly not understand why their new files don't appear. >>>>>>>>> >>>>>>>>> That sound strange. We'll have a look at it. >>>>>>>> >>>>>>>> Ross, >>>>>>>> >>>>>>>> Our FVT team just tried that out and reported that it works as >>>>>>>> expected. >>>>>>>> >>>>>>>> Alan >>>>>>>> _______________________________________________ >>>>>>>> cifs-discuss mailing list >>>>>>>> cifs-discuss@opensolaris.org >>>>>>>> http://mail.opensolaris.org/mailman/listinfo/cifs-discuss >>>> >>>> _______________________________________________ >>>> cifs-discuss mailing list >>>> cifs-discuss@opensolaris.org >>>> http://mail.opensolaris.org/mailman/listinfo/cifs-discuss >>>> >>> I think this issue is very time related, I set the time wrong on >>> purpose on the CIFS server, and I saw this issue very frequently on >>> CIFS clients. >>> Synced the clock again to NTP, and haven't been able to reproduce it >>> again; however even with the time grossly incorrect, the issue was >>> still intermittent. >>> >>> I'll post again if I see the issue with correct time. >>> >>> -- >>> Brent Jones >>> br...@servuhome.net >>> > > _______________________________________________ cifs-discuss mailing list cifs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/cifs-discuss
