I’m a ‘computer guy’ and at times have been the ‘tech guy’ for departments I’ve worked for, but when it comes to anything regarding the network, I always call the IT people. Who knows, maybe I’d be able to solve the problem, but if I didn’t set the network up or spend time getting intimately familiar with it, I figure I’d better leave well enough alone. I’ll probably waste my time and potentially screw things up.
Certainly there are times when one’s judgment about one’s own ability vs the ability of any IT folks who may or may not be helpful needs to come into play regarding fixing tech problems. However, unless the tech person really knows you, I see no reason why he or she should trust your judgment in that regard. Especially due to the fact that if you screw something up, it’s his or her time that gets spent putting it right.
Not necessarily, but it’s Bad Practice to start flipping switches if you do not know what you’re doing.
For one thing, it has the tech troubleshooting a network that’s no longer in the state it was when it failed. Useful information is lost - routers (well, better models) save useful statistics and logs in memory, and much of this information is irretrievably lost when the switch is flipped.
In (mercifully) rare circumstances, things can go really, really bad.
For instance, power-cycling and disconnecting cables in the wrong sequence can throw a switched network into something called “VTP meltdown” - the details are a tad on the esoteric side, but the effects are hour-long outages and the sort of stuff that alters careers.
Or the network may be in trouble due to a condition known as “route flap”, where the routers spend too much time telling each other what goes where and not enough time actually moving data. If the immediate reaction is for everyone to go out and turn the routers on and off, it actually worsens the core problem - and it sure as hell makes life harder on anyone trying to figure out what the heck is up.
It’s a shotgun approach. Sometimes it works, there’s no denying that. Sometimes it makes things much worse. (FWIW, I’m posting this from a Network Operations Centre controlling just over 500 Cisco routers on several continents. Our local techs do not power-cycle without authorization.)
And we will continue to, if no one takes the time to just explain things to anyone.
I can see getting upset about someone walking in and rebooting the router. What I cant see is why, since you have someone who is actually willing to get off their ass and take some initiative, the anger lasts more than a few seconds. All that need be done is say something like:
"Hey, I appreciate that you want to help, but in the future, please dont just do that. "
Since the most common (recurring) network problems in any company are pretty straightforward and have pretty clear symptoms, you could then go on to explain to her that if this happens this way, check this; if this happens that way, check that; if you see this message, do that; and if the problem is not like any of those, then page someone in IT or Ops (since its unclear whether the OP is desktop support or network ops).
I also wonder why there was/is no lock on the hardware room door, and if I was the OP, Id put one in.
I dont think the OP meant it, but in the OP a little and more in others posts, there is a certain amount of the union attitude ‘Thats not my/your job’. Christ, the lady at least didnt say ‘Oh the networks down, Im off to starbucks, see ya!’, she stayed and tried to fix the problem so she could continue working. If she doesnt know what shes doing, then try to show her; initiative like that is rare, expecially among cube drones.
IT-guy ringing in here : the first thing I would do (and most Network Administrators) would be to power-cycle the network equipment.
This solves most problems and should never cause a problem, if properly configured.
If I was at home and a user would call me, I would tell them to power-cycle the network equipment as well.
If that doesn’t help I would try and analyse the problem further.
I think a lot of IT-personnel should indeed get out of their ivory tower and start acting like people.
IT guy here. I think the real reason for the rage is “It’s been down since yesterday.” One of my biggest pet peeves is that, while we have a documented proceedure for putting in tickets here, nobody will use it. They will wait for days or weeks until they see me in the halls.
That, and some routers don’t have on/off switches, so you have to plug and unplug them. Not to mention that they tend to be tucked away places, so people messing with them have this unholy tendency to pull out patch cables.
Frankly, it’s not so much the switching on and off issue… heck, I might even tell the person to do it… as the doing it without informing IT that you’re doing it. Or that there’s a problem. Because then secondary problems appear… and you never quite catch the end.
I caught that as well; but having worked at quite a few tech companies, its equally likely that either the ticket wasnt submitted because of laziness or that the ticket wasnt submitted because that particular IT dept has a reputation of not handling tickets ASAP. Ive worked in places where tickets were handled as soon as they came in, and Ive worked in places where tickets werent even looked at until lunch/smoke break/bathroom break/Half Life Tourney were done. Not using the escalation system can either be a sign of laziness, or a sign of experience as to how quickly things get fixed once escalated. Im convinced that some IT depts only have escalation systems so that they can finish whatever RPG theyre playing before dealing with a problem, rather than be interrupted by someone popping their head into the cube and blowing a perfectly good Unreal 2 match just to fix something as lame and unimportant as, oh, the email server being down.
It depends on the size and complexity of the network. Power cycling is fine for my home network of 2 computers and a printer, and it’s probably fine for a small business network, but it’s clearly not fine for Spiny Norman’s vast international network.
First of all, shame on Greenback for not having a secure network. You should have the network room locked at all times. As well as any network racks. This is like #1.
Then again 9 out of 10 support personnel are dumb as shit. You simply do not just go and reboot network devices without any knowledge of what you’re doing. Even if you do, it is the responsibility of the Network Admins, not yours. So stay the fuck away from them. “It was a reboot of a couple machines, fuckstick” lol, WTF is this?
It still doesn’t matter. If it isn’t your job, stay the fuck out of it and open a ticket. I know my way around networking devices, but I don’t go around rebooting the network devices, even if I know what they do. This is a security no-no! I would so recommend you be fired or suspended for fucking with something I oversee and am responsible for.
Most people with the type of knowledge your talking about consists of maybe building a PC, and installing a slipstreamed version of XP on their computers. Whoopty-FUCKIN-DOOOOO!! Maybe they even installed a linksys router and cable modem in their home. CCNA certified I say! I’m so sorry, but this doesn’t even come close to the networking education required to make the decision to reboot a corporate router!
Sure, it may have worked sometimes. But you are not qualified to make this decision on your own. You should only do this if a Network Admin instructs you to do so. Never, NEVER do it alone! Guessing with shit like this, uninstructed would have you fired so fast! And deservedly so!
While its a good idea, its not always possible. Corporations often don’t have secure space for such equipment.
Otherwise, I’d have to agree with you, ParentalAdvisory though you were a bit harsh. Rebooting the router on most systems - unless you’re talking about really big corporate networks, is a good idea usually. if you’ve got nothing but end users, reboot, wait ten minutes, and you’re probably good.
Of course, that doesn’t excuse Miss Reboots-a-lot there. She should not have been fiddling with the equipment like that. At the least she might have damaged the conectivity and jarred some cables; at worst she would could have messed up several servers and screwed up athe local mailbox, and judging from the “I’m doing overtime for this mess” comment, it was probably somewhere in the middle.
Of course, I might be biased. If I ever hear another idiot come tell me the nwtwork is down while I’m happily surfing. I heard that 10 times a week for 4 years. It was correct a few times, sure, but only because the Uni had a crappy network department I had to work with.
Yeah, but even for some small companys, if there is room for some routers and switches, there is room for a small locked rack or cabinet. I’m not saying you have to have a room with a door, racks are sufficient (locked ones). They come in all sizes.
I may have been a little harsh. I could’ve been nicer. Put it this way, you’re not going to perform heart bypass surgery if you’ve read the book 100 times and are sure you can do it, right? Leave it to those who’s job it is to perform, even if you know how to fix it. You may have a deadline on your accounting project on your local network, you may even be Cisco master. If you cannot get someone to fix it, or instruct you to, it’s just too bad, your boss cannot fault you for it if it’s not done because the network was down for 2 days.
Now if you’re a small company without network support, I guess that’s kinda up in the air. But if you’ve got a guy who’s sole repsonsility is the network, you need to live it up to him.
There are four key things here that, in my mind, make this a little above the normal ‘this user is just trying to be helpful, give the guy a break’ thing.
The first one is, from the OP, this is the 4th time this particular person has gone back there and messed around with things. Its not mentioned specifically, but based by how upset the OP is, its likely he has mentioned his desire for her to stay away from those machines before.
The second is, as mentioned by others, if you are having an issue the goal isn’t always necessarily to just ‘fix it’ as soon as possible. Its better to see it while it is broken to better understand whats going wrong. Rebooting might be a quick fix, but understanding what the root cause is will be better for you in the long run. You can’t do that if someone has already gone in and yanked the plug on the thing.
Third, from the message that this person left saying “I went in and tried turning things on and off in random order and it didnt work” this is not a person who has the first idea of what those routers do and how they are configured. This was obviously a ‘push buttons and see if something happens’ approach. It’s not being egotistical to say this person doesn’t know what they are doing and therefore shouldn’t be doing it.
and, lastly, since it was down all afternoon and there was no call put in, it seems likely this wasn’t a life or death kind of emergency. I could almost see going in there and trying to fix it on your own if you HAVE to have that network up or payroll isn’t going to go out. But if its a small enough deal you can wait until the next day before you even mention it to someone, there’s no need for that person to have waited until the expert was available.
Like I said to metacom, our network is a pit thread unto itseelf. Our router and waverider in that office are in a back storage room but there is no lock whatsoever. Heck, our head office servers are sitting out in the open on a counter. It’s terrible. However, the fact that they are exposed does not give anyone the right to start their own unsupervised troubleshooting session.
The target of this thread knows enough about computers to be dangeous, but she thinks she knows enough to be helpful.
Guinastasia - The Waverider is provided by the local ISP and is basically a wireless connection to their network.
…this is what most bad Network Administrators would do, because in the event that the router was not the problem (as in the OP), you’ve just interrupted network service for a bunch of people who were not having any trouble. If all anyone wanted to do was browse the web, then there’s no problem, but if (for example) someone was in the middle of a large data transfer, you’ve just made a lazy attempt at sweeping a problem under the rug and broken something else in the process.
Couple of ~days~?? Dude, if youre working at a company that let the network be down for two days and you werent fired after about the first 4-6 hours, youd better hang on to that job.
And its not a matter of being worried as to whether the boss is going to be pissed; this might come as a suprise, but some people actually give a shit about getting things done that they are being paid to get done. If Im developing something and the network goes down and I cant work, IT or OPs had better be on it within a half hour, and if they estimate its going to take something along the lines of two hours, then two hours later it had better be done, or phone calls are going to upper management explaining that they are paying me triple digits to sit on my ass.
As far as two days goes, that is plain fucking absurd; Id walk in and fix the fucking thing myself, or call upper management and offer my wifes contracting services - who holds the three main Cisco certs - and suggest to them that they replace the worthless idiots on their Ops/IT staff. Two ~days~!! That is absolutely unfucking believably not acceptable. 48 hours straight working to get a network up? Youve got worse problems than just hardware. Or even worse, you arent actually talking about going home at any time before the thing is back up are you? Incredible. Two days for a network to be down spoken of like it was acceptable, and people wonder why shit gets outsourced.
Whatever. As I said, IT/Ops teams are split between very professional ones, and very immature incompetant ones. Much of the vitriol here is probably from people who have experience in the latter.
I work weekends and see sites go down all the time, 2 days is not uncommon for a network to be down. This is especially true for remote plants on the weekends. If however a user does need to get work done, and if it’s business critical, then yes, a Sev 1 is issued and work is performed. However, if you’re working on something that can actually wait 2 days and is not, again, business critical, yes, it will wait 2 days. A determination is usually made right then and there with the user and network support. You’d be suprised how many times a developer is screaming that they have to get their work done, only to get on the phone with the Network guy to find out it can actually wait a week :rolleyes:. It’s a money issue, and it is very real one in the business world. You should still not however try to fix the problem yourself, or invite your wife to do it. Or maybe I’m just an immature and incompetant IT person. :rolleyes:
