smokris's questions

I'm running a SmartOS-based ZFS NAS, with a bunch of SATA disks directly connected to the logic board (no fancy disk controller / backplane or anything).

A 1.5 TB disk in my zpool recently failed. iostat says:

# iostat -En c2t5d0
c2t5d0           Soft Errors: 0 Hard Errors: 8 Transport Errors: 21 
Vendor: ATA      Product: ST31500341AS     Revision: SD17 Serial No: [...]
Size: 1500.30GB <1500301910016 bytes>
Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 
Illegal Request: 89040 Predictive Failure Analysis: 0 

I hotswapped the disk with a new 2.0 TB disk, yet iostat still says the same thing — Size: 1500.30GB.

How can I tell the system to rescan the disk so it identifies the correct disk size? (I'd like it to probe the correct size so I can eventually expand my zpool.)

Already tried

1. devfsadm -c disk as suggested on a similar question — no effect.
2. cfgadm -c configure sata0/5::dsk/c2t5d0 — it just says Device already configured; no effect.

The SmartOS Global Zone defaults to the UTC timezone. I'd like to schedule some cron jobs, and I'd like them to automatically follow daylight saving time changes, without requiring me to manually change the UTC offset every time the daylight saving change occurs.

How can I change the timezone for my SmartOS Global Zone?

I'd like to run munin-node on my SmartOS Global Zone, to remotely monitor hardware and software status.

I've tried pkgin in munin-node-1.4.5, but installation fails because it can't create the munin user account (which makes sense since /etc/passwd is part of the root ramdisk).

What's the right way to do this?

In Mac OS Server 10.6, when I select a user account, uncheck the "access account" checkbox, and click Save, I get this error message:

Error of type eDSAuthFailed (-14090) on line 3912 of /SourceCache/WorkgroupManager/WorkgroupManager-361.3.1/Plugins/UserAccounts/UserAdvancedPluginView.mm

It happen for all user accounts I've tried. I'm able to make changes to other attributes (e.g., passwords); I just can't disable accounts.

Any ideas about how to fix this, or to work around it in order to disable user accounts?

Already Tried

• changeip -checkhostname says "The names match. There is nothing to change."

I'm running a Windows Server 2003 guest instance in Xen 3.x. This DomU runs fine for a day or two, then stops responding — I don't get any network response, and I can no longer connect to Xen's VNC console for this DomU.

xm list shows this:

Name                                        ID   Mem VCPUs      State   Time(s)
Domain-0                                     0  6508     8     r----- 1161159.4
[A working Linux DomU]                       1   512     1     -b----  68711.1
[The hung Windows DomU]                      5   512     1     ------  67234.2
[Another working Linux DomU]                 3   512     1     -b---- 163036.4

(What does the ------ mean? The xm manual explains what each of the six states mean, but not what no-state means.)

If I xm destroy and then xm create the Windows DomU again, it boots right back up (with the Windows alert The previous system shutdown at [...] was unexpected.), and then stops responding after another day or two. Aside from that alert, there's nothing relevant in the Windows event log. Also, I'm using Munin to monitor disk, network, process count, CPU usage, and memory usage; the Munin graphs don't indicate any resource exhaustion or other suspicious activity prior to the hang.

I checked /var/log/xen/*.log, but no log messages are generated at the time the server stops responding.

How should I proceed in troubleshooting this?

I'm running Mac OS 10.6 server, and have a set of 10.6 and 10.7 workstations connected to it for authentication and MCX. Users have network profiles (home folder stored on server, via AFP).

When users are logged in, they can change their desktop background. So far so good.

However, the next time they log in, their desktop background has been reset to "Andromeda Galaxy.jpg".

Though MCX is enabled and used to control other settings, MCX is disabled for the Desktop.

What is keeping the users' desktop background from being preserved? How can I fix it?

I'd like to make use of LDAP groups in my Subversion Edge "Repository Access Rules".

For example, if LDAP user smokris is a member of LDAP group dev, I'd like to be able to grant the dev group access to test-repository, without having to explicitly add each member of group dev to the Repository Access Rules.

What's the syntax for that?

I tried:

[test-repository:/]
dev=rw

[/]
*=

...but user smokris is denied access.

I have puppet master running on a Xen dom0, and 3 domUs syncing to it via an hourly crontab puppet agent --test.

About 80% of the time, the puppet agent --test completes successfully:

info: Retrieving plugin
info: Caching catalog for test3
info: Applying configuration version '1333319732'
notice: Finished catalog run in 5.08 seconds

The other 20% of the time, it fails midway, with errors such as the following:

err: Could not retrieve catalog from remote server: Error 400 on SERVER:
     Could not find class iptables for test1 at /etc/puppet/manifests/site.pp:1 on node test1
warning: Not using cache on failed catalog
err: Could not retrieve catalog; skipping run

or

info: Retrieving plugin
info: Caching catalog for test2
info: Applying configuration version '1333319732'
notice: Finished catalog run in 24.73 seconds
err: Could not send report: Error 500 on SERVER: 

Internal Server Error

private method `gsub' called for WEBrick::HTTPStatus::RequestTimeout:Class

 WEBrick/1.3.1 (Ruby/1.8.5/2006-08-25) OpenSSL/0.9.8e-rhel5 at
 puppet:8140

or

info: Retrieving plugin
err: Could not retrieve catalog from remote server: execution expired
warning: Not using cache on failed catalog
err: Could not retrieve catalog; skipping run

or

info: Retrieving plugin
info: Caching catalog for test3
info: Applying configuration version '1333319732'
notice: Finished catalog run in 9.47 seconds
err: Could not send report: Error 408 on SERVER: Request Timeout 

During this time, I've not made any changes to the Puppet configuration — it just sporadically fails.

I'm running puppet-2.7.12 on CentOS, and followed the setup instructions described on http://docs.puppetlabs.com/learning/agent_master_basic.html.

Any ideas about how I can troubleshoot this?

I have a Nexenta system that's been running fine for the last few years. Recently one of the system disks failed --- the syspool was part of a ZFS mirror.

If I configure the system's BIOS to boot from the working drive, the text GRUB ("GRUB" followed by a space) appears on the screen, and the system doesn't respond.

If I boot from a Nexenta installation CDROM and try to set the root as follows:

root (hd1,0,a)

...I get the following message from GRUB:

Filesystem type unknown, partition type 0xbf

However if I were to issue the above root command on a fresh Nexenta installation, I would see:

Filesystem type zfs, partition type 0xbf

How can I get GRUB to recognize the surviving root mirror disk, so I can boot from it? And, in the future, what procedure should I follow when a device in a root mirror fails, so that I can properly recover (and boot from any drive in the mirror pool)?

What factors should I consider when deciding whether to use Apache's mod_perl or mod_cgi, when configuring an existing web application?

I'm trying to copy files on a SMB fileshare. When I attempt to copy the files in Finder, I get the following error:

 The operation can’t be completed because you don’t have permission to access some of the items.

Copying via Terminal.app (using a simple cp command) works just fine.

Permissions on the folders (as seen from the computer attached to the fileshare) are as follows:

Source:

 dr-xr-x---   2 smokris  staff    16384 Oct 13 10:55 .
 dr-xr-x---@ 61 smokris  staff    16384 Oct 13 10:56 ..
 -r--r-----   1 smokris  staff    53970 Oct 13 10:55 ._IMG_3823.JPG
 -r--r-----@  1 smokris  staff  3135600 Oct 13 10:55 IMG_3823.JPG

Destination:

 drwxrwx---  2 smokris  staff    16384 Apr  9 10:17 .
 drwxrwx---  3 smokris  staff    16384 Apr  9 10:15 ..

Any ideas?

I've got a working NexentaCore Platform 2.0rc2 (OpenSolaris b104) machine, which I'm trying to bring up-to-date using apt-clone upgrade --- NCP 3.0a4 (OpenSolaris b124).

The upgrade process appears to complete successfully, however the machine reboots right after I select the new checkpoint in the GRUB menu.

When I boot the new checkpoint with "-v", I see the following messages right before the screen flickers out for reboot:

 WARNING: failed to resolve 'scsa,probe' driver alias, defaulting to 'nulldriver'
 WARNING: failed to resolve 'scsa,nodev' driver alias, defaulting to 'nulldriver'

Not sure if this is related.

Any suggestions about how I can troubleshoot this?

In a generic, modern unix environment (say, GNU/Linux, GNU/Solaris, or Mac OS X), is there a good way to determine which mountpoint and filesystem-type a particular absolute file path is on?

I suppose I could execute the mount command and manually parse the output of that and string-compare it with my file path, but before I do that I'm wondering if there's a more elegant way.

I'm developing a BASH script that makes use of extended attributes, and want to make it Do The Right Thing (to the small extent that it is possible) for a variety of filesystems and host environments.

I'm administering a cluster of Macs, with shared login. (The shared login thing wasn't my idea, but I've got to live with it.)

We're running into problems where users connect to fileshares and click the "Remember this password in my keychain" checkbox, which presents a security problem.

Is there some way to disable this option, so that users can never store their fileshare credentials?

I'm administering a number of Macs, whose users would like to be able to make use of iCal.app (and hence iPhone sync) to view and edit Oracle Calendar events.

Is there a good way to bidirectionally interface Oracle Calendar's SyncML with iCal.app?

Or alternatively is there a way to bridge a SyncML server to provide CalDAV service (which iCal.app can natively handle)?