861 Words. Plan about 4 minute(s) to read this.
These are facts I did not know (or was unsure of) that I picked up while going through the CD-ROM testing engine for chapters 1-6. Take this with a grain of salt, as these testing engines are notorious for technical errors.
- Ethernet V2 frames have a 2-byte type field. 802.3 frames have length and LLC field not present in V2.
- SNMP “Inform” messages elicit a “response” message.
- MST uses RSTP. Duh. I got a question about MST convergence, and I couldn’t remember what timers MST used, as I spent very little time reviewing MST. But…it uses RSTP. So convergence in MST would be 3x hello by default (or 6 seconds), after which it will begin forwarding out the alternate root port immediately.
- Grr. I’m going to have to write down my IP subnets to keep from making STUPID STUPID STUPID mistakes. I had to do a summary route, and I was actually correct in my head (to a point), but screwed up because I wasn’t thinking clearly about the bottom boundary of one of the subnets. If I’d been paying better attention, that would not have happened.
- I blew I/G and U/L bits in an ethernet MAC. I remember what they were and even where they were, but then my mind froze trying to remember if given “03” as the first 2 digits in a MAC, whether the first byte was “0” or “03”. Silly mistake, “03” being correct, both U/L is most significant, I/G is second most, thus “3” implying that both bits were on (00000011).
- 802.3ab is gig over copper. I remembered 802.3z was gig over fiber, and ruled out the rest, but it wasn’t like I was 100% sure of the answer.
- Hey, stupid! “NVRAM” means non-volatile RAM. That’d be different than regular RAM there, genius-boy. And yeah, VTP clients don’t store VLAN information in NVRAM. Only, when I read NVRAM, my brain said “yeah, that’s the stuff that goes away when you shut the power off.” Silly brain. I want a new one.
- VTP information isn’t issued out a non-trunk port. Hmm. I have to check that one.
- VLAN 1001 is prunable. VLAN 1 is not.
- A short prefix length does not mean a small subnet. Actually, it’s the opposite: the shorter the prefix length, the larger the subnet. Getting to read these things naturally will help with some question confusion.
- If an interface fails, STP will go about the business of choosing a new root port immediately. STP doesn’t have to wait for maxage to expire.
- I learned that if you don’t read the question…the WHOLE question…you will screw yourself for absolutely no good reason. All the phrases in the questions mean important things. Don’t assume something you think to be obvious, especially in a scenario question. Read the question carefully, digesting each bit thoroughly. In the case of one question, I overlooked the fact that the switches had NOT been cabled before the config was applied, and it made all the difference in the right answer.
- I learned that ambiguous questions really irritate me. Certain questions about who picks which port in an STP topology can’t be determined unless you know who’s root bridge and therefore, where the hellos are coming from.
- LACP supports a concept of standby ports. You can have 8 active and 8 standby ports. Huh. You can’t do that with PAgP.
- The TCP window field is advertising a window for traffic flowing the opposite direction, but had nothing to do with traffic flowing in the current direction. Sort of like “hey, if you decide to send me some data at some point, here’s my window.”
- By sheer luck (well, I eliminated a couple of bogus answers), I guessed that if the source and destination SAP are set to 0xAA on an IEEE 802.3 frame, the frame format in use is IEEE 802.3 SNAP.
- I’m reminded that sometimes questions will overwhelm you with irrelevant information (diagrams and IOS configs) to try to throw you off the scent of a straightforward question.
- NTP stratum of 7 is the default.
- I need to look at what commands are required on a router interface when configuring it as a dot1q trunk. It appears you are required to configure a native VLAN, and I’m not sure I remember that exactly.
- I need to brush up a bit on ISL.
- I really hate it when a choice appears obviously correct, but the test engine says it’s wrong and doesn’t explain why. That’s called “lazy” on the part of the guy who wrote the question. I’m okay with being wrong – but tell me why I’m wrong so that I learn from my faux pas and don’t repeat the blunder.
- I need to review how DP’s are chosen on a segment, but my brain is getting fuzzy. I’ve been at this for over 2 hours, trying to get through this question set.
- LoopGuard doesn’t react to a BPDU? Hmm….I have to look that one up.
- Well, 76% on the first run through…55 out of 72 questions in that batch. Not too bad, but definitely needs some work. I’m not too far off the mark, though. I think my study method is okay. So I’ll soldier on to Chapter 7 tomorrow.
Ethan Banks writes & podcasts about IT, new media, and personal tech.
about | subscribe | @ecbanks