Ramblings

Looks like I have three packages on the way.

Well, back on Monday I thought I had my complete order from Newegg on the way. Well, I was wrong. (I know, that doesn’t happen very often :)

To place the order I used my father’s credit card. He’s funding this project (as he has most of them [Gollum, Frodo, parts in Bilbo] in the past) with no strings attached. Using someone else’s credit card is where my issues started …

I had to place the order in two parts because one of the items I needed two of (the PCI IDE controllers) had a limit of one per customer (basically per order unless you seriously abuse it and then they will crack down). The order that had only a single IDE controller was about $45 and shipped almost immediately. The cost wasn’t high enough for Newegg to do any address verification stuff.

However, the second (actually the first one I placed) order was significantly more than $45. After shipping charges it was approaching $1,000 which was well over the limit that triggers address verification. So, since my address did not match the billing address for the card, the order was put on hold and I was emailed a note telling me I needed to contact Newegg directly to resolve the issue.

Phone Call #1 (2/14/2005): No real suprise here. I was told that I needed to either change the shipping address to the billing address or have the card owner (my dad in this case) add my address to the list of valid shipping addresses for his card. No big deal as I was able to call dad and get him to do that almost immediately.

About 2 hours pass with no email back from Newegg like I was told to expect once my address had been added.

Phone Call #2 (2/14/2005): They looked up my order and asked why I was calling back (they noted each phone call on the order). After explaining that I never recieved the email telling me that everything was OK, I felt the need to check in and make sure that everything was going ahead. I was politely told that everything was fine, but that they needed the contact information on the card holder. No problem … I gave it to them. I was then informed that I should recieve the confirmation email within about two hours. OK, thanks … click.

No email before 5PM, but since they are two hours behind me I figured I’d wait until I got home later that evening (I hung around on campus for a while then headed home). I got home about 6:30 or 7:00 and still hadn’t recieved anything. Since I figured they were wrapping up for the day, I’d just call the next morning.

Phone Call #3 (2/15/2005): I had checked their website before calling and noticed that they had La Porte misspelled (la parte) and figured that may have caused some problem. So, I talked to yet another person at Newegg about my woes. Apparently the bank had rejected the address because Newegg sent them my name with my parent’s address. Wrong! So she fixed that (and the spelling of La Porte) and told me to wait the 1 – 2 hours for the confirmation email.

By the time that I had a chance to follow up on the day’s activities, it was already about 3PM and I still hadn’t recieved the email. So …

Phone Call #4 (2/15/2005): What’s going on? Oh, the bank was closed by the time we got around to trying the information you gave us earlier today (call logs are nice). Man! This is getting rather annoying. I like Newegg, but this was getting out of control. I was told that they would process it as soon as the bank reopened.

And that brings us to this morning. I checked the website … my order was still on hold and I hadn’t recieved any emails indicating what was going on. So …

Phone Call #5 (2/16/2005): Why is my order still on hold? Oh, hang on a second … … … OK, I just ran your order through and everything looks good. (About this time I recieved an email from Newegg letting me know that my order passed the address verification and was on it’s way to getting packaged.) After I made sure that it was indeed proceeding normally again, I hung up. A few minutes later their website updated and now shows that it has made it to step 3 of 5 (packaging and shipping are left).

So it finally appears that my parts are heading this way (well, after packaging).

Well, I decided to give it a whirl. I have installed WordPress version 1.5. I haven’t bothered to get my old theme (Blue-Fade) working yet, and I’m not sure if I will. One of my plugins (adds links in the html header for site navigation) decided to misbehave on subpages, but it worked just fine on the front page. Go figure … I may attempt to debug it later, but it isn’t mine so I’m not scrambling to fix it.

Well, the Gollum supplement/replacement has been ordered from newegg. It’s on the way … more about it later since it’s class time right now.

We are looking into building a new machine to act as our server. Gollum (the current machine) has done great recently, but we are looking at this more from a redundancy viewpoint than a strictly replacement one.

Here is a wishlist at newegg showing the parts that I have thrown together for it so far. Nothing is written in stone …

The idea is to build a new machine that stays here in Longview with me. It would become the primary machine once I have it set up and running. Gollum would then get moved to La Porte and sit on the cable connection there and handle secondary DNS, and failover everything (or at least as much as we can figure out)

Hopefully, if, for whatever reason, my connection here in Longview dies (thank you Longview Cable, you suck!) then Gollum in La Porte would be able to handle the load for a while. At a minimum I’d like Gollum (in a backup role) to be able to handle web, email, and dns. The web stuff doesn’t have to be bleeding edge current and the email would hopefully just queue up for ultimate delivery to the new machine (and maybe provide access to the email via POP3 or IMAP).

Any thoughts? I still need to really sit and think about how I want this to work. Oh, and John Spiegel has offered to host backups of actual files (nightly backups are really needed and I’m not doing it right now) and a slave DNS setup for my domains.

Wow. I went digging through my usage stats for the website today (as I do fairly regularly) and saw the list of search strings that have lead users to my site. It’s quite interesting that someone searched for “bert cox dallas” and hit my site. I hope they were searching for the same person that I know … it must have keyed off the link I have on the right side to his blog. My favorite search string so far this month, though, is “sample reprimand letters for church officials”. What on earth did I write that someone found by searching for that!

As an aside, dad’s site transferred over 50GB last month. Most of which seemed to be in the last 10 days or so. I can’t say I didn’t notice … it kept poor Gollum rather busy for that period. It wasn’t a high load average, and in fact barely touched the processor, but it was the constant maxing of my outbound connection that I noticed :) I like to give him a hard time about it, but it really doesn’t bother me and if it did bother me, there are ways to “fix” it.

I guess I forgot to mention that I got email back up first thing the next morning. It was a permissions problem … it happens everytime I upgrade vpopmail. It changes the user on the files to something other than it should be, so suddenly it can’t write to the folders/files. Once I remembered what happened last time it was an easy fix. I also had to update the config that allows me to run spamassassin and clamav on the emails. The upgrade dropped a new version of the integration script into production, but didn’t merge any of my config options. Oh well.

I decided it was long due … I began the process of running updates on Gollum (my main server). The updates were going smooth … until all of the sudden Gollum just FROZE. So, I switched my monitor over to the server and what do I find? A kernel panic message. Great!

So I reboot. No big deal, other than the fact that I was enjoying the fact that I had over 180 days of uptime. Well, the reboot was a big deal. Apparently either the kernel panic or some action on my part currupted the patition table on one of my drives in the RAID. Ok, so several hours later I found out what to do (there were other things that caused that time to pass so quickly that were not related to Gollum). So, I have the array up and running again. I don’t know what I would have done without the array.

Anyways, as a result of the upgrades, email broke again (every time I upgrade!) so that is going to have to wait until tomorrow morning from the office :)

I just got logrotate installed on Gollum this afternoon. I must say, it’s not done like I would have expected. I would have expected some daemon that runs in the background and wakes up to do it’s work, but instead you have to add it to the crontab. I guess that makes sense too, but Gentoo doesn’t say anything about that at installation. I set up the messages log, apache logs files, and mail logs to be rotated out weekly, but it’s supposed to keep 12 weeks of compressed logs around before finally emailing me the oldest as it’s rotated out.

Gollum had a little hickup this morning as many of you noticed. For some reason my cable modem wouldn’t renew it’s DHCP lease. When the lease died, so did my connection. Thankfully when I reset the modem it just started working again and I kept my old IP. I had to make a trip to my apartment for lunch just do fix everything, but it wasn’t so bad afterall. I may just start going back there everyday.

Again, it’s been a week since I posted last and I apologize for that deficiency. A lot has happened since that last scream for help. Sometime shortly after posting the edit on the 11th, my cable connection died. I was messing with my router remotely when it happened, so I just assumed I broke something that could be fixed when I got home that evening. Man was I ever wrong. I got home, reset all my hardware (not the computers), and give it another whirl. The first thing I noticed was that my router picked up a new IP address and the second was that there was only one DNS server listed where there is normally two.

It took me a second to realize that the new IP address that was assigned to me was a 10.0.X address. In other words, it was well within the private network range of addresses. It all suddenly made sense, they had cut me off. After some quick poking around I realized that all outgoing DNS requests (for stuff like http) were getting resolved to the same internal IP address. Just to show you how messed up their system is, I know the desired outcome was to redirect me to a page telling me that my service had been discontinued ‘automatically’. However, their http server (IIS I think) was not configured to ignore the hostname that was requested. As a result, the page would not load.

After looking up the IP address that everything was being redirected to, I typed it in myself. Ya, just a simple ‘you account has been disabled due to non-payment’ page. I was current on my bill, I didn’t owe them anything. Well, a very long story somewhat short is that I spent the next day on the phone. I dealt with about four people at Longview Cable and Cablelynx before I got things all fixed. Thanks goodness I finally was connected to someone who knew what was going on.

In other news, the craziness has subsided some around these parts. I got one of the major projects out the door and working (the one I was freaking out about the errors last time) by the end of last week and I got the second one for University Advancement up and running yesterday. The one for Advancement was actually rather simple, but I needed the time to work on it which I didn’t get until yesterday.

Now it’s just back to Traditional Web Registration, a large enough project on it’s own.