Patch Name: PHCO_18317 Patch Description: s700_800 10.20 mount(1M) cumulative patch Creation Date: 99/04/29 Post Date: 99/05/07 Hardware Platforms - OS Releases: s700: 10.20 s800: 10.20 Products: N/A Filesets: OS-Core.UX-CORE Automatic Reboot?: No Status: General Superseded Critical: No Path Name: /hp-ux_patches/s700_800/10.X/PHCO_18317 Symptoms: PHCO_18317: If there are networking problems, the mount(1M) of a local file system may take longer to complete than is acceptable. PHCO_17699: If a file system is in /etc/mnttab but is not really mounted, perhaps due to a bad or out-of-date /etc/mnttab file, the mount command fails with the message that the file system is already mounted, even though the file system is not mounted. PHCO_16873: Concurrent mounts/unmounts may cause mount(1M)/umount(1M) to fail with: "mount: file system table may be corrupt". PHCO_16235: The NFS mount option -oretry=n will not work in all cases with the 10.20 ACE 2 bundle. PHCO_16043: The NFS mount option -oretry=n will not work with the 10.20 ACE 2 bundle. Only one attempt to mount the file system will be made regardless of the number of retries requested. PHCO_14175: The NFS mount options -obg and -oretry=n do not work as described when the mount command is used, with the exception of the 'mount -a' case. There are no retries performed if the initial mount fails, and thus, the subsequent retries cannot be run in the background. PHCO_13847: This patch is part of the 10.20 ACE 2 bundle which adds networking enhancements to 10.20. New networking features supported in ACE 2 include NFS Version 3.0, AutoFS, and CacheFS. PHCO_9696: When mount is invoked without any arguments, the dates presented at the mount times for each file system will show the hour as if the user is in time zone EST5EDT, regardless of what time zone the user is actually using. PHCO_8008: When /sbin/mount was called with file system types not included in HP-UX (such as by other applications), and the first item to mount was a directory, mount will fail with an invalid file system type message. Defect Description: PHCO_18317: If there are networking problems, the mount(1M) of a local file system may take longer to complete than is acceptable. The problem is that gethostbyname() is called unnecessarily for many file systems when trying to determine if a file system is already mounted. This is usually not a problem, but if DNS problems are being experienced, this may result in a longer delay than needed when trying to mount local file systems. Resolution: The fix is to change mount so that only file systems with a ":" in the name are treated as remote. All other file system types are considered local, and will avoid the gethostbyname() call, thus restoring acceptable performance. PHCO_17699: The mount command relies on /etc/mnttab to determine if a file system is already mounted. Therefore, if the /etc/mnttab file is bad or out-of-date due to, for example, an unexpected reboot which does not update /etc/mnttab when it unmounts file systems, the mount command fails with the message that the file system is already mounted, even though the file system may not actually be mounted. This can have serious ramifications in certain situations, such as when a reboot happens at system init time such that a critical file system such as /stand cannot be mounted. Resolution: When the mount command finds the file system in /etc/mnttab, it then checks the kernel file system table to see if the file system is, in fact, mounted. If it IS mounted, it appropriately fails with an "already mounted" message. If it is NOT mounted, this indicates that /etc/mnttab and the kernel file system table are out of sync, so it updates /etc/mnttab and then proceeds with mounting the file system. PHCO_16873: Concurrent mounts/unmounts may cause mount(1M)/umount(1M) to fail with: "mount: file system table may be corrupt". Resolution: The libc routines that access /etc/mnttab need to change so that they can better ensure the integrity of /etc/mnttab, and also can help prevent deadlocks. In cooperation with libc changes, the mount(1M) command needs some integrity checks when it reads /etc/mnttab, to help compensate for not having read locking. This prevents mount(1M) from using an out-of-date /etc/mnttab file, or from using a corrupt or empty file and failing with, "mount: file system table may be corrupt". In addition, mount(1M) needs to change the way it updates /etc/mnttab, so that write-locking is minimized. All these changes help reduce the possibility of a deadlock situation, especially in HA environments, where multiple processes may be attempting to access and/or update /etc/mnttab at the same time. PHCO_16235: The NFS mount option -oretry=n will not work in all cases with the 10.20 ACE 2 bundle. In these cases only one attempt will be made. PHCO_16043: If the user uses the "-oretry=2" option or any larger number, only one attempt will be made to mount the NFS file system if the 10.20 ACE 2 bundle is installed. PHCO_14175: If the user uses the "-oretry=2" option or any larger number, only one attempt will be made to mount the NFS file system. Since the "-obg" option relies on retrying the mount if it fails, this will not appear to work since "bg" puts the mount in the background on subsequent attempts and currently only one attempt is made. PHCO_13847: New functionality to support networking features in 10.20. PHCO_9696: Use the command 'mount' to show the times when file systems were mounted. The hour will be relative to time zone EST5EDT regardless of the time zone the user has specified in his environment. PHCO_8008: When the first argument to mount is a directory, the mount wrapper would try to process it with its known file system types, and when this did not fit into the known categories, would fail. The mount wrapper should not be checking this value, and should let the file system specific version of the mount command determine if it is an error. SR: 5003456723 1653293282 1653270181 5003430850 5003419911 5003382309 4701378117 5003344622 4701330431 Patch Files: /sbin/mount what(1) Output: /sbin/mount: $Revision: 78.9.1.9 $ libcteam_nightly_dav,981025_20:00 libc.a_ID@@/main/r 10dav/libc_dav/libc_dav_cpe/cup_libcteam_nig htly_dav/1 /ux/core/libs/libc/archive_pa1/libc.a_ID Oct 25 1998 20:08:07 PATCH_10_20: mount.o 99/04/29 cksum(1) Output: 2056121742 266240 /sbin/mount Patch Conflicts: None Patch Dependencies: None Hardware Dependencies: None Other Dependencies: None Supersedes: PHCO_8008 PHCO_9696 PHCO_13847 PHCO_14175 PHCO_16043 PHCO_16235 PHCO_16873 PHCO_17699 Equivalent Patches: PHCO_17698: s700: 11.00 s800: 11.00 Patch Package Size: 320 KBytes Installation Instructions: Please review all instructions and the Hewlett-Packard SupportLine User Guide or your Hewlett-Packard support terms and conditions for precautions, scope of license, restrictions, and, limitation of liability and warranties, before installing this patch. ------------------------------------------------------------ 1. Back up your system before installing a patch. 2. Login as root. 3. Copy the patch to the /tmp directory. 4. Move to the /tmp directory and unshar the patch: cd /tmp sh PHCO_18317 5a. For a standalone system, run swinstall to install the patch: swinstall -x autoreboot=true -x match_target=true \ -s /tmp/PHCO_18317.depot 5b. For a homogeneous NFS Diskless cluster run swcluster on the server to install the patch on the server and the clients: swcluster -i -b This will invoke swcluster in the interactive mode and force all clients to be shut down. WARNING: All cluster clients must be shut down prior to the patch installation. Installing the patch while the clients are booted is unsupported and can lead to serious problems. The swcluster command will invoke an swinstall session in which you must specify: alternate root path - default is /export/shared_root/OS_700 source depot path - /tmp/PHCO_18317.depot To complete the installation, select the patch by choosing "Actions -> Match What Target Has" and then "Actions -> Install" from the Menubar. 5c. For a heterogeneous NFS Diskless cluster: - run swinstall on the server as in step 5a to install the patch on the cluster server. - run swcluster on the server as in step 5b to install the patch on the cluster clients. By default swinstall will archive the original software in /var/adm/sw/patch/PHCO_18317. If you do not wish to retain a copy of the original software, you can create an empty file named /var/adm/sw/patch/PATCH_NOSAVE. Warning: If this file exists when a patch is installed, the patch cannot be deinstalled. Please be careful when using this feature. It is recommended that you move the PHCO_18317.text file to /var/adm/sw/patch for future reference. To put this patch on a magnetic tape and install from the tape drive, use the command: dd if=/tmp/PHCO_18317.depot of=/dev/rmt/0m bs=2k Special Installation Instructions: None