Devnull's questions

I have a cache device that fell out of the zfs array because of errors. Im pretty sure it was something i did - messing with cabling or something. It checks out fine, plus its just a cache device. When i try to readd it, it fails with the following error:

# zpool add array cache /dev/sdj2
cannot add to 'array': one or more vdevs refer to the same device

is there a way to remove the reference from the zpool array so i can readd it? ive tried remove:

# zpool remove array /dev/sdj2
cannot remove /dev/sdj2: no such device in pool

Troubleshooting an issue with an FTP connection going through an iptables firewall and seeing some strange issues with passive mode.

We are trying to connect to an FTP server and get the directory listing, and it seems to work in PASV mode in all cases, but times out when EPSV is sent instead. The server understand EPSV, because that works with iptables disabled.

tcpdump shows the client sending the EPSV request, and then the server does not respond after EPSV is sent. if the client sends PASV, everything works as expected.

nf_conntrack, nf_conntrack_ftp, nf_nat, and nf_nat_ftp are all loaded and the appropriate rules are in place.

When attempting to change ISOs on a powered on guest instance, set-cddrive is returning an error but completing the command. A couple different commands produce the same result:

$theVM | Get-CDDrive | Set-CDDrive -NoMedia -Confirm:$false
$theVM | Get-CDDrive | Set-CDDrive -Connected $true -ISOPath $iso -Confirm:$false

both produce: Set-CDDrive Operation is not valid due to the current state of the object

What state might that be? Is set-cddrive dependent on either CusomizeVM_Task or ReconfigVM_Task?

The kicker here is that the operation appears to succeed, even though it returns an error.

Edit:

error has now changed to "ide0:0 already exists" without any changes to scripting.

Found this KB, indicating it might be a bug: http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2092716

In one user mailbox anything sent from any distribution list in our domain shows up as being from 'System Administrator'.

It only happens when they are viewing the inbox using outlook (OWA is not affected), and its persisted across computers (though it did not happen immediately). When other users view the inbox from their outlook install (ie open users folder), everything appears as normal.

Other folders are not affected. if a message is moved into a subfolder, the sender displays properly.

Because of the persistence, and it only affecting one user, I suspect some user behavior is causing this, but i cannot determine what. Ive checked the contact list, and its not that.

How does a routing and remote access server pick up routes to advertise to clients, and where can a list of routes being provided be reviewed?

One of our users is reporting that the server is advertising a route which is valid, but has never been entered in the static routes section of the RRAS server, or told/provided to them at any time. In fact the static route section is currently empty.

RIP is not configured at this time, so its not coming from there that I can tell.

we are using PPTP, not l2tp.

I need to delegate permissions to edit user information (phone numbers, address, etc) in AD. Its a 2003 domain. Ive already delegated: 'Read all user information' and 'Read, Write, Read and write general information, Read and write phone and mail options, read and write public information, read and write personal information.

So far none of these have helped the specific user edit details. Its been logged off and back on already too.

What am i missing?

Reading over (and then testing out on a test database) the setup for mirroring it appears that one is supposed to leave the mirror database in restoration mode for normal operations.

is this correct? at whatever point it becomes necessary to initiate the failover (im not using a witness), will the primary database be put into restoration mode? this seems goofy.

We have a server 2008 r2 box with a 60gb c:\ partition that over time completely fills up. The culprit appears to be winsxs, which is a well documented issue without any resolution beyond microsoft jamming their fingers in their ears and repeating 'its not a real issue'. This is all accurate from an extremely narrow viewpoint, however explorer, symantec liveupdate, wsus, and a host of other applications that live on this machine disagree and refuse to operate properly.

Ive seen where people are displeased because winsxs takes a few more gig than it should, however this has literally filled the entire disk. Is this an artifact of wsus, or some other crazy behavior? Anyone know anything i can do to stop (and hopefully reverse) this usage? The other machine we have running 2k8r2 is doing just fine, so its not a general OS issue, that i can tell.

i have a cisco asa5520 in transparent mode. the interface is on the same subnet as some windows machines, which are generating a lot of broadcast traffic that is filling up the logs.

is there any way to have it not log that its blocking those packets? its a bunch of these messages: "through-the-device packet to from management-only network is denied: udp src..."

im also seeing some of those zeroconf requests that id like to drop logging for.

i tried to just put a rule on the management interface, but apparently thats not allowed.

it looks like turning the syslog filter up to error might solve both issues, but will i be missing any potentially life saving messages by ignoring the info bits?

On a windows 2k8R2 DC box the eventlog service wont start. When i try to start the service the machine returns: "Windows could not start the Windows Event Log service on Local Computer. Error 13: The data is invalid."

event viewer wont open either.

obviously, i dont have an event log for this error.

ive already tried clearing the event logs. that didnt do anything.

the other DC isnt experiencing this issue.

Does there exist a router that supports multiple GRE connections over NAT? Im currently running pfSense, and it only supports 1 at a time. I understand why, its just a drag as there are multiple people in this office trying connect to the same VPN server.

Obviously there are other ways to solve this, like different VPN setups, multiple interfaces, etc, but for a variety of reasons Id much prefer multiple GRE connections.

if one has a run of singlemode fiber terminated in a patch panel, how would you then connect it over to a device? singlemode patch? it was my understanding that singlemode had minimum refraction distance requirements.

can anyone recommend (or recommend to avoid) a source for asset labels?

id like foil, potentially with a color logo if possible. so if youve had something like that printed before, please speak up.

google returns far too many results to be manageable.

This has me beat...

we have microsoft dynamics with a sharepoint business portal interface using sharepoint services 3.0. im trying to configure the sharepoint interface to accept basic auth, so that reverse proxying works. ive found a couple of places that need changes, both in sharepoint and IIS, but whenever i switch it over, certain pieces of business portal start to error out - things like the project communicator and expense reports section. basically, it looks like anything that appears in an IFRAME doesnt work.

while we dont use these pieces yet, we will, so this is basically a showstopper for me.

so far ive tried:

• just changing IIS from integrated to basic
• changing IIS and the sharepoint config for the site to basic
• configuring the reverse proxy to present authentication the way that IIS expects with integrated auth enabled (ie <domain>\<user>)

im hesitant to try bringing up another duplicate site (a best practice i read in a sharepoint book) to support both integrated and basic, but i dont fully understand how sharepoint works, and im loathe to break it again. additionally, since i cant make the one we have work the way i want, im skeptical ill be able to do so for an additional site.

anyone have any magic that might help me out of this?

edited to include the error below:
Error:

Connector:Unspecified HTTP error. HRESULT=0x800A1518 - Client:An unanticipated error occurred during the processing of this request. HRESULT=0x800A1518 - Client:Sending the Soap message failed or no recognizable response was received HRESULT=0x800A1518 - Client:Unspecified client error. HRESULT=0x800A1518

faultcode=Client

faultstring=Connector:Unspecified HTTP error.

faultactor=

detail=Connector:Unspecified HTTP error. HRESULT=0x800A1518 - Client:An unanticipated error occurred during the processing of this request. HRESULT=0x800A1518 - Client:Sending the Soap message failed or no recognizable response was received HRESULT=0x800A1518 - Client:Unspecified client error. HRESULT=0x800A1518

We recently had a full power outage and had to shut down basically all devices in the server room.

Being a smaller org, we still have some very small business practices im working to change over time, like having a bunch of external disks in production use. In this case, all those disks needed to be powered down, but some of them were connected to the primary rack PDUs and had to be unplugged manually, and then replugged later. What would have been very nice would be to have some/most of these types of devices on their on plugstrip for easy control.

Does anyone know of plugstrips (bonus points for ones that do a good job managing wallwarts) that have covered switches, to prevent accidental switching?

in putting together a new installation, one of the requirements is that we use fiber to the desktop. we are going to use multimode, but im not sure which connector.

is there a material (ie cost) difference between SC and LC? what about durability?

It would seem to me that the fragile part (the fiber) doesnt benefit much from either connector, so is either OK?

any other advice or things that seem obvious in hindsight would also be appreciated.

edited: also, if anyone has some insight as to which 1000baseSX nics one should avoid (if any) that would be most appreciated.

When using apache to reverse proxy, is it possible to forward http://somesite.com/ and http://somesite.com/foo to different servers? Say one goes to http://internalserver1/ and the other goes to http://internalserver2/foo

I can obviously just drop further down the path (ie http://somesite/bar) but its not desirable.

Ive been looking at monitoring for a while. My org didnt have any before i came other than 'whered my yahoo go'. It appears that most packages out there focus on negative monitoring (ie, this service/host was up and now its not). This seems like a valid first step, but what can you look at past that for positive monitoring (ie that port wasnt up, and now it is, or hey look thats a new DHCP host)? I suppose its possible to have a declaration for every single port/network address in nagios, but that seems cumbersome.

Does anyone know of a better tool for monitoring ports/hosts for affirmatively down?

We have several users connected to our exchange server using activesync using a variety of devices. The documentation for mobileadmin appears to indicate that the app should provide a list of active devices somewhere, but i cant seem to find it. Is there a config change i need to make, or are the docs inaccurate?

anyone know of a tool to inspect and compare BKF files? or compare BKF contents with whats live on the filesystem?

lets say i have a known-good BKF, and all it contains are the files i want to compare. i have another BKF that is known-bad. how could i tell the differences between them (files, permissions, contents, etc)?

one solution would be to restore to different locations and look at them that way, but being able to compare the two BKFs would be much cleaner.