Tag Archives: asking questions

My trouble with bonded interfaces

In an effort to improve the redundancy of our network, I have all of our blade servers configured to have bonded network interfaces. Bonding the interfaces in linux means that eth0 and eth1 form together like Voltron into bond0, an interface that can be “high availability”, meaning if one physical port (or the device it is plugged into) dies, the other can take over.

Because I wanted to eliminate a single point of failure, I used two switches:

The switches are tied together to make sure traffic on one switch hits the other if necessary.

Here is my problem, though: I have had an array of interesting traffic patterns from my hosts. Some times they’ll have occasional intermittent loss of connectivity, sometimes they’ll have regular time periods of non-connectivity (both of which I’ve solved by changing the bonding method), and most recently, I’ve had the very irritating problem of a host connecting perfectly fine to anything on the local subnet, but remote traffic experiences heavy traffic loss. To fix the problem, all I have to do is unplug one of the network cables.

I’ve got the machine set up in bonding mode 0. According to the documents, mode 0 is:

Round-robin policy: Transmit packets in sequential
order from the first available slave through the
last. This mode provides load balancing and fault

It would be at least logical if I lost 50% of the packets. Two interfaces, one malfunctioning, half the packets. But no, it’s more like 70% of the packets getting lost, and I haven’t managed to figure it out yet.

If you check my twitter feed for yesterday, I was whining about forgetting a jacket. This is because I was hanging out in the colocation running tests. ‘tcpdump’ shows that the packets are actually being sent. Only occasional responses are received, though, unless the other host is local, in which case everything is fine.

There are several hosts configured identically to this one, however this is the only one displaying this issue. Normally I’d suspect the firewall, but there isn’t anything in the configuration that would single out this machine, and the arp tables check out everywhere. I’m confused, but I haven’t given up yet. I’ll let you know if I figure it out, and in the mean time, if you’ve got suggestions, I’m open to them.

Tape Labeling in AMANDA

Alright, I’m once again appealing to people with more experience than I have.

Waist deep in learning AMANDA, I have questions about tape labeling. I’m attempting to find the best practices for labeling tapes, both physically (via barcode stickers on the tape) and the “logical” label written using the ‘amlabel’ command.

Anyone who knows better, please correct the following overview: For anyone who doesn’t know (this included me 2 hours ago), AMANDA uses a program called ‘amlabel’ to write a label to the beginning of the tape so that it always knows which tape is inserted. It keeps records of every tape that it ever writes, so that in the future when you attempt to recover data, you can insert the specific tape that it asks for.

Since this electronic label can be in nearly any format, many people that I’ve seen commenting in various places have the knee-jerk reaction to label their tapes “Monday”, “Tuesday”, etc etc. Many of the old-timers discourage that, because if a tape runs over, you’ve screwed up the labeling system. Many people suggest “Daily1, Daily2, Daily3”, etc etc.

My question is this: What are your tapes labeled, and do your AMANDA labels match the physical labels on your tapes? Also, if you have a barcode reader, do you order customized labels to take advantage of that, so that your barcode matches your AMANDA label, or do you have some sort of mapping between the two in a spreadsheet (or AMANDA itself)?

By the way, if you’re looking for customized labels, I found that https://www.labelarchitect.com has some very configurable options. Color, # of characters, etc. Feel free to share your sources if you’ve got some suggestions!

How to ask questions on the internet

This was originally written on October 23rd, 2007 and post in my livejournal. It ends short, but the idea isn’t too bad, and the content isn’t yet out of date.

As the internet has become ubiquitous over the past decade, it has become our primary research tool. Seemingly endless in the knowledge it contains, we use an array of tools to mine for data, most of them from Google.

Sometimes, our search ends up short. The infinite monkeys on infinite typewriters have yet to produce the particular work we need, and it’s up to us to prod them into movement. We must ask a question to the faceless, and hope someone succumbs to our petitioning.

I’m here to help you figure out how to do that.

You’d think it would be easy. You’d think you could just ask what you wanted to know. Folly.

At some point during your research, you must have come across two of the most common instances where someone asked a question and received absolutely no help at all. Allow me to illustrate:

Our goal: Why the foo widget doesn’t appear to work when the bar widget is installed

Obvious (i.e. wrong) question:
I’ve got foo widget, but it doesn’t work when bar widget is installed. Any idea why?

While to the non-reptilian brain, this might appear to be a perfectly valid question, if you submitted this to a forum, you would most likely be greeted with derision and mocking. Why? You didn’t give enough information.

Suppose you asked the question and received several such replies. You might be tempted to say “screw you guys”, find another forum, and ask the same question with more detail. Not a bad idea, but you definitely need to be careful. Here’s why:

More detailed (i.e. wrong) question:
All: I’ve got foo widget version x.y.z running on my slackbuntuhat 7 machine. When I install bar widget 37.9 I am getting the following errors occuring in syslog.
– cut 30 lines of text –
I’ve read through the sourcecode and at the point it generates this error, the $arrBaz looks like it might be overflowing, but I can’t tell if it’s a bug or a clever programmer. I modified the source, and fixed the original error, but now I am getting this output:
– cut 40 lines of text –
Anyone else had this problem?

The response:

*crickets chirping*

Alright, you have corrected the initial flaw, a dearth of information. In it’s place, you have inserted more information than anyone other than the actual developers of the software are likely to know anything about. Unless you’re on a listserve where the authors are frequent posters, you’re out of luck.

The proper way is to tease support out of the other commenters. You must coax information out of them a bit at a time, just whetting their curiosity. Maybe start like this:

All, I’m having issues with foo widget x.y interacting with bar widget 37. foo works fine, but when I install bar, it goes belly up. Anyone seen this happen?

After that, you’ll probably get a reply asking for what platform it’s on, log output, etc etc. Include the information they request, and always end the post with a request for more help. That way, anyone casually browsing by who has the knowledge will see your post being the last, with no reply, and they might respond. If you overwhelm them with information in any one post, you’ll get dropped like a hot potato.

Thanks for reading, and good luck.