what did you learn today? (part 2)

[ramases]

K0DE[/url]":2pdvvcco]For GUIs I have the background on the prod boxes bright red but I am not aware of a way to custom color the filer's SSH sessions (not that that's an excuse but I save my own ass a lot of the time, would have been nice in this case).

You may not be able to color the background, but as long as you have a bash you can always colorize your prompt and text output via .bashrc.

CanSpice[/url]":2pdvvcco]
I've been advocating for focus-follows-eyes and the more technologically-advanced focus-follows-brain for years now.

This. In the mean time, I make do with mouse-focus with auto-raise and auto-dim.

 

[afidel]

K0DE[/url]":ho7tle2m]I don't think anyone noticed. Well, I am sure they noticed, I don't think anyone thought it was actually down, I had it offline for ~60 seconds where it was actually unreachable and it was towards the end of lunch so our least busy time of the working day.

Still

For GUIs I have the background on the prod boxes bright red but I am not aware of a way to custom color the filer's SSH sessions (not that that's an excuse but I save my own ass a lot of the time, would have been nice in this case).

You can use custom session profiles in putty, set the default background color for each filer.

 

[Gandalf007]

You can use ANSI colors in your bash prompt (PS1) -- many distros do this, e.g. root's prompt is red.

 

[sryan2k1]

the filer's

Sounds like it's an embedded system (VNX, etc) and not a normal linux machine. It may not be possible to change the bash prompt.

 

[BitPoet]

sryan2k1[/url]":30crr7xi]

the filer's

Sounds like it's an embedded system (VNX, etc) and not a normal linux machine. It may not be possible to change the bash prompt.

IIRC NetApp runs a BSD or Linux variant at the bottom somewhere. VNX block storage is embedded windows (or was last time I looked) VNX NAS Control Station is linux, but the data movers are a custom OS, and I don't remember any .server_config commands to change the text colors, but I don't think they have ssh anyway so it's a moot point.

 

[techix]

TIL what happens when an unsuspecting coworker plugs a HP DL580 G7 into the (clearly marked) 400 Hz power distribution block. (Loud pop, smoke, & profanity).

 

[afidel]

techix[/url]":3jf5aofp]TIL what happens when an unsuspecting coworker plugs a HP DL580 G7 into the (clearly marked) 400 Hz power distribution block. (Loud pop, smoke, & profanity).

Ouch, that's an expensive beast to fry that way, couldn't he have learned that mistake on something a bit cheaper? Also why do you still have 400Hz power? Do you have a really, really ancient IBM box about somewhere?

 

[techix]

afidel[/url]":38jl6mvm]

techix[/url]":38jl6mvm]TIL what happens when an unsuspecting coworker plugs a HP DL580 G7 into the (clearly marked) 400 Hz power distribution block. (Loud pop, smoke, & profanity).

Ouch, that's an expensive beast to fry that way, couldn't he have learned that mistake on something a bit cheaper? Also why do you still have 400Hz power? Do you have a really, really ancient IBM box about somewhere?

Would have been nice to fry something cheaper- hopefully it is only the power supply.

As for 400Hz- the lab used to test parts for these(now it is just extra space for labbing):

http://en.wikipedia.org/wiki/UGM-133_Trident_II

 

[w00key]

Does anyone know how long it takes for those Intel DC S3500s to hit the shelves in Europe? One vendor I called today (.nl) said 3-4 weeks, which is kinda meh. The price was really good tho, € 321 for 300GB incl 21% sales tax, that's just slightly more expensive than consumer level drives and way below what the 320 used to be (~€1.5/GB).

 

[M. Jones]

BitPoet[/url]":3u6vi7jo]

sryan2k1[/url]":3u6vi7jo]

the filer's

Sounds like it's an embedded system (VNX, etc) and not a normal linux machine. It may not be possible to change the bash prompt.

IIRC NetApp runs a BSD or Linux variant at the bottom somewhere.

NetApps are NetBSD, but as with a lot of other vendors NetApp will give you a song and dance about how there's no NetBSD in ONTAP any longer because they rewrote everything.

One problem with customization is the risk of becoming reliant on customization. You're not far away from making a mistake because your prompt wasn't the expected red. It's probably better to make a habit of being cautious.

 

[sryan2k1]

Also why do you still have 400Hz power?

400Hz power is pretty standard in anything that flies.

 

[kennedye]

Not so much learned, as dodged a bullet; I managed to arrange for us to go with ClamAV for virus-scanning our (Unix) PCI systems as opposed to having to install Symantec Endpoint Protection (shudder).

 

[Rick25]

That MS fixed the issue with Mailbox cleanup after a move (SP2 RU3 issue) in SP3 RU1

http://support.microsoft.com/kb/2763065

Scorp, just a note the description for that KB in the summary of RU1 is "Move request log is logged when you move a mailbox in an Exchange Server 2010 SP2 environment"
but when you go into the link it's (correctly) "Mailbox move operation fails when Update Rollup 3 or a later update rollup for Exchange Server 2010 SP2 is installed"

 

[Entegy]

If you use a disclaimer transport rule, do not install SP3 RU1. It will cause your HTML-formatted messages to be caught up in the poison queue as the transport process crashes over and over.

 

[Rick25]

Yep, thanks for the heads up. We're not caught by that but I was just reading the comments on the Ehlo page for RU1 that some people are running into issues with opening additional mailboxes through OWA after RU1. Going to have to do some testing. Oh they also mentioned no more RU releases for SP2, which is fair enough I guess.

 

[sryan2k1]

ASA's doing RSA via SDI should not have more than one RSA server defined, doing so will cause random authentication timeouts.

You only specify which server you'd like to be the primary on the ASA (either the master or any replica) and the RSA servers send down a list of alternates during the node secret exchange.

 

[M. Jones]

afidel[/url]":309dfhto]

M. Jones[/url]":309dfhto]I've discovered a twelve year old Red Hat Linux instance running a highly-customised RDBMS replication service, in revenue production. The hardware is probably twelve years old too, having 384 megabytes RAM, but I haven't yet tracked down the physical machine.

What's the uptime!?

Six days.

Any more archaeology like this and I'm changing my login to 'Indiana Jones'.

 

[afidel]

sryan2k1[/url]":3lruiwd3]ASA's doing RSA via SDI should not have more than one RSA server defined, doing so will cause random authentication timeouts.

You only specify which server you'd like to be the primary on the ASA (either the master or any replica) and the RSA servers send down a list of alternates during the node secret exchange.

And what happens if the 'primary' is down when your ASA's come up?

 

[sryan2k1]

It's not an issue, the list gets downloaded and stored in the ASA after the first successful auth / node secret exchange. After that only one needs to be available.

 

[afidel]

M. Jones[/url]":1cea3no6]

afidel[/url]":1cea3no6]

M. Jones[/url]":1cea3no6]I've discovered a twelve year old Red Hat Linux instance running a highly-customised RDBMS replication service, in revenue production. The hardware is probably twelve years old too, having 384 megabytes RAM, but I haven't yet tracked down the physical machine.

What's the uptime!?

Six days.

Any more archaeology like this and I'm changing my login to 'Indiana Jones'.

Oh, I was expecting like 10 years since they last patched, maybe it was setup by a bad windows admin and they thought it needed weekly reboots?

 

[Barmaglot]

TIL that Microsoft Security Essentials installed on Windows Server 2008 R2 (VM under Xen) can mess it up to the point where the virtual disk service is unavailable, and iSCSI LUNs won't mount. Fortunately I wasn't the one who installed it, but I was pulling what little is left of my hair out for hours until I stumbled upon MSE being the culprit.

 

[afidel]

Barmaglot[/url]":7u52p7a2]TIL that Microsoft Security Essentials installed on Windows Server 2008 R2 (VM under Xen) can mess it up to the point where the virtual disk service is unavailable, and iSCSI LUNs won't mount. Fortunately I wasn't the one who installed it, but I was pulling what little is left of my hair out for hours until I stumbled upon MSE being the culprit.

Huh, wonder what it was, since MSSE and FEP are the same core engine there shouldn't have been a problem.

 

[Danger Mouse]

afidel[/url]":2k27xebf]

Barmaglot[/url]":2k27xebf]TIL that Microsoft Security Essentials installed on Windows Server 2008 R2 (VM under Xen) can mess it up to the point where the virtual disk service is unavailable, and iSCSI LUNs won't mount. Fortunately I wasn't the one who installed it, but I was pulling what little is left of my hair out for hours until I stumbled upon MSE being the culprit.

Huh, wonder what it was, since MSSE and FEP are the same core engine there shouldn't have been a problem.

My guess

 

[Barmaglot]

Well, I didn't have a chance to get rid of Xen for that particular customer (not that I ever could, for reasons unrelated), but uninstalling MSE fixed it right quick.

 

[ronelson]

You can use ANSI colors in your bash prompt (PS1) -- many distros do this, e.g. root's prompt is red.

At the very least, PS1 should have the hostname in it. Course, one still has to read it, the problem is usually typing/pasting a command in and hitting enter without thinking at all, in which case colors and names matter for naught.


Also, another vote for never take my fucking focus away ever you brain dead programmer. So many horrible programs, from installers to video games to expensive business software likes to steal focus, repeatedly. My favorite are the ones that reset your mouse cursor, even if you disable the OS level setting to do so. Yay!

 

[Big Wooly Mammoth]

Symantec Endpoint Protection's scan takes focus away several times a day; fucking pisses me off.

Our attorneys are the only other people who seem to notice.

 

[Entegy]

IS THAT WHAT THAT IS!?

Holy shit, it's been bugging me so much I thought I had malware, but I end my current job in a week, so I never bothered to reformat the thing. Glad I didn't now! I kept noticing that my current window would just randomly lose focus.

That might also explain why my computer randomly gets slow as if all my apps are paged out around lunchtime. I didn't even think it could have been SEP. But of course, it's the shitty Symantec product.

 

[Technarch]

I actually didn't mind that about SEP, because when it decides to run a scan, it takes all the CPU and RAM as well as focus, so I'm not going to get anything done anyway. In some respects it's good that it takes focus, because it saves me a click when I go to cancel the scan.

What was really fun was having SEP and FireAMP stepping on each other all the time, shouting that they'd found eicar.txt in each other's quarantine folder. That went on for a couple of months before I figured out how to get SEP to play nice.

 

[jshiplett]

FreeNAS is stupid easy to get up and running (and supports VAAI!). If I had known this, I probably would have ditched OpenIndiana for my lab.

 

[nshank]

TIL that Forefront TMG needs to die. In a de-compiler.

Also, IPv6 firewalling is difficult due to limitations of requirements (non-legacy adapters on Hyper-V. Don't ask.)

But, on the plus side, my SCCM/SCEP deployment is done. And thus far, every thing is working. Yeah, I'm afraid to.

 

[Barmaglot]

nshank[/url]":2xwnawnq]And thus far, every thing is working.

File under: famous last words.

 

[nshank]

Barmaglot[/url]":34qen94u]

nshank[/url]":34qen94u]And thus far, every thing is working.

File under: famous last words.

Eh, the Earth is going to spin off it's axis in the next day or two anyway, so what does it matter?

/sarcasm

 

[Barmaglot]

Is it just me, or did GoDaddy just go down completely?

 

[SandyTech]

Barmaglot[/url]":32zoee5t]Is it just me, or did GoDaddy just go down completely?

Dunno about completely, but some of the stuff we host on GoDaddy (thankfully internal stuff) has stopped responding...

 

[Barmaglot]

All the domains we host there (which is quite a few) are not responding, and even their main website is down, so it appears they've once again fucked up their DNS.

Also, in unrelated news, our <expletive deleted> beloved lawmakers just announced yet another <expletive deleted> DST change. What I wouldn't give, to have them, all 120 of them, go over all the affected systems and adjust the settings, by hand. This would actually be very useful - it'd keep them busy and unable to cause any more harm. <expletive deleted>

 

[hernias]

Barmaglot[/url]":3uaz9bix]All the domains we host there (which is quite a few) are not responding, and even their main website is down, so it appears they've once again fucked up their DNS.

Also, in unrelated news, our <expletive deleted> beloved lawmakers just announced yet another <expletive deleted> DST change. What I wouldn't give, to have them, all 120 of them, go over all the affected systems and adjust the settings, by hand. This would actually be very useful - it'd keep them busy and unable to cause any more harm. <expletive deleted>

DST has no place in modern society. Either make it year-round or eliminate it entirely.

 

[Barmaglot]

Tell that to our penguin friends.

 

[Danger Mouse]

...that the DataDomain DD660 I have, required not only to have the data retention settings changed in Commvault, but a disable/enable of the CIFS backup target (D2D2T) filesystem in the DDOS in order for it to mark the aged segments as aged and available for "cleaning" (purging).

Previously, I got back about 54GB during a clean. now I got back 508GB and it's not finished yet.

It's also possible that it's a bug in the OS or during the multiple upgrades to get to current (just upgraded to 5.2.2.4).

Thank goodness it's fixed, because I need to have at least ONE clean backup prior to our data center shutdown/move on July 4th.

Yep, that's right. My manager scheduled our move for July 4th and through the weekend. If it was Christmas, I'd be calling him Scrooge Every single person who was told about the schedule said more or less "WTF".

 

[PaveHawk-]

I have to use the Dell Online Solutions Configurator. Within it, when configuring a server, is an option for cooling that reads: "Fresh Air Cooling".

What is that? Do my servers that I order get a few free RUs in a rack near an ocean breeze? Or do they get to be settled out on a nice little garden somewhere?

 

[theevilsharpie]

More adventures in XenServer land!

So my customer is retiring a number of old servers, one of them being a DNS server. As part of the decomissioning process, I'm changing the IP settings of some of the hard-coded hosts to point to the IP of the new DNS server. Sounds simple, right?

:bigdumbgrin:

Here's the command to change the DNS settings on a XenServer host:

xe pif-reconfigure-ip uuid=<pif_uuid> mode=static DNS=<new_dns_ip>

Oh, and here's a little note from Citrix's documentation:

Caution should be used when changing the IP address of a server, and other networking parameters. Depending upon the network topology and the change being made, connections to network storage may be lost. If this happens the storage must be replugged using the Repair Storage function in XenCenter, or the pbd-plug command using the CLI. For this reason, it may be advisable to migrate VMs away from the server before changing its IP configuration.

(BTW, XenServer's iSCSI storage support is hilariously bad, but that's a topic for another time.)

I picked a lightly-used XenServer host in the pool as a testbed and put it into maintenance mode, which went smoothly. Then I ran the command, and..... nothing happened. The command appeared to succeed, but xsconsole and /etc/resolve.conf still showed the old DNS server entries. A quick Google search on how to change a DNS server address in XenServer showed that I'm performing the correct procedure, but in order for the change to take effect, the XenServer host has to be rebooted. Really?



I wasn't about to disrupt the health of the XenServer pool by rebooting hosts, so I figured I'd push this off until the next downtime window (the old DNS server will still be operational until then). I tried to take the host out of maintenance mode, and received the following error:

This operation cannot be performed because HA would no longer be guaranteed for this pool. To perform this operation, you must disable or reconfigure HA.

To put this message in perspective, consider the following:

1. I have so much spare memory capacity in my XenServer hosts that a single host can run all of my VMs, and still have 25% of its memory available
2. I have multiple hosts in this pool that are not in maintenance mode, have plenty of memory for additional VMs, and are capable of performing live migrations
3. I'm adding capacity to my pool, not removing anything
4. This server was just in the pool, hosting VMs

The solution? I had to disable HA, take the host out of maintenance mode, restore the VMs it was running, and re-enable HA again. Apparently, Citrix doesn't quite understand what HA is supposed to do.

:facepalm:

I cannot wait to get away from this sorry excuse of a hypervisor.