Display card miner failure drags switch
1. Power failure
to do a good job of external power supply, generally through the introction of independent power line to provide independent power supply, and add voltage regulator to avoid instantaneous high voltage or low voltage phenomenon
if conditions permit, uninterruptible power supply can be added to ensure the normal power supply of the switch. Some provide voltage stabilizing function, while some do not. Pay attention to the selection. Professional lightning protection measures are set in the computer room to avoid the damage of lightning to the switch
In practical use, cable failure often leads to switch system or port can not work normally, so this kind of failure is also classified as switch hardware failure. For example, the connector connection is not tight, the cable proction sequence is wrong or not standardized, the cable connection should use the cross line but use the straight line, the two optical fibers in the optical cable are interleaved, the wrong line connection leads to the network loop, etc 
3, mole failure
switch is composed of many moles, such as stack mole, management mole (control mole) and expansion mole. The failure probability of these moles is very small, but once there is a problem, it will suffer huge economic losses. If you are not careful when plugging in and out the mole, or you are collided when carrying the switch, or the power supply is unstable, it may lead to this kind of failure
4. Port failure
if the optical fiber plug is accidentally soiled, the optical fiber port may be polluted and cannot communicate normally. Many people like hot plug connector, theoretically speaking, it is OK, but this also unintentionally increases the failure rate of the port
in general, a port failure is that one or more ports are damaged. Therefore, after troubleshooting the computer connected to the port, we can judge whether it is damaged by replacing the connected port. In case of such failure, you can try to clean the port with alcohol cotton ball after the power is turned off. If the port is damaged, you can only replace it
5. Backplane fault
under the condition of normal external power supply, if each internal mole of the switch can not work normally, the backplane may be broken. In this case, even the electrical maintenance engineer may have no way to do, and the only way is to replace the backplane< br />
As the core "hub" of data exchange and transmission in LAN, the operation performance of switch directly affects the efficiency of network data transmission. In general, it is not easy for ordinary layer-2 switches to fail. However, recently, an ordinary layer-2 switch in the computer room of the unit training center encountered an inexplicable fault. Considering that the fault is a bit strange and easy to occur, it is not easy to solve, Therefore, this paper will contribute the troubleshooting process to communicate with you, to help you improve the efficiency of switch troubleshooting
fault playback
there are 36 ordinary workstations and 2 special examination servers in the training center room of the unit, which are connected together through three switches to form a medium-sized training and examination network, in which a Cisco ws-c2950sx-24 switch is used as the main switch of the local area network, In addition, two d-linkdes-1024d switches are used as the secondary switches of the LAN. Each secondary switch is connected to 18 ordinary workstations at the same time, and each secondary switch is connected to the main switch of the LAN. Two examination servers are directly connected to the main switch of the LAN, All computers in the LAN are connected to the external network through the main switch (as shown in Figure 1)
recently, a part of ordinary workstations in the computer room of the training center suddenly can't access the special server for examination, while another part of workstations can normally access the contents of the server. The author rushed to the wiring cabinet of the computer room and carefully checked the network connection equipment, It is found that all the signal lights in the control panel of a certain switch are all on without flashing. It is obvious that all the ordinary workstations connected to the switch will naturally have the phenomenon of network failure
troubleshooting
at first, I thought that the switch cache overflowed, which led to the unexpected paralysis of the switch, so I did not hesitate to cut off the power supply of the switch, and after a while, I switched on the switch again. After the switch started stably, I specially tried to access the examination server in the LAN from the ordinary workstation, The results show that the normal workstation has been able to access the server. Seeing that the fault has been successfully solved so soon, the author is secretly happy and complacent about his network management level; But who ever thought that the front foot just left the computer room of the training center, the back foot received the "help" from the teacher of the training center, saying that the network fault that had just been solved appeared again
when I return to the connection cabinet of the computer room again, I can see that the switch with previous problems seems to be in a state of paralysis. Is this the disaster caused by the network virus in the LAN? Although the author knows that there must be one or another virus program in the common workstation used for teaching and training in the LAN, if it is really caused by the virus program, another secondary switch with the same model and connection mode should also have the same fault (as shown in Figure 2), But now another secondary switch works normally, and all workstations connected to this switch can also access the server normally; According to this comparison, the author almost concludes that the switch paralysis is caused by the previous switch's own hardware failure
as the switch hardware faults mainly include the faults of port, power supply, mole, backplane and other components, the author checks the faulty switch one by one from these aspects. When eliminating the port factor, the author thinks that at most one or several ports may be damaged, and it is impossible that all ports in the switch will be damaged at a certain time. Therefore, the author infers that the switch hardware fault should not be at the port location; When checking the power supply, the author found that the fault switch and another secondary switch are connected to the same power socket, that is, the power input of the two switches is completely the same, but now one switch can work normally, which indicates that the power input should not be a problem. After eliminating the port and power factors, the author estimates that the hardware failure of the switch is likely to occur in the mole or backplane components, and for this kind of hardware failure, the only thing we ordinary users can do is to replace the switch
because there is no other free switch on hand, the author has to temporarily disconnect the power supply of the faulty switch and prepare to report to the leader before purchasing a new switch. While those workstations can not access the Internet, the author intends to "strangle" all workstations, so that they can welcome the arrival of the new switch with a new look. After a long period of virus killing, the author found that there are many viruses in the network. By chance, the author saw one of them who is familiar with the virus. When I looked up the relevant information of the virus on the Internet, I found that the virus can cause a broadcast storm; So I suddenly came up with such an idea in my heart. Could it be the broadcast storm that caused the frequent paralysis of the faulty switch? Now that the virus has been completely removed, will the faulty switch still be paralyzed? With the mentality of having a try, the author reconnects the power supply of the faulty switch. At the beginning, the operation state of the switch is very normal, but the author is worried that the switch will continue to be "paralyzed" after a long time, so he specially stayed beside the switch for more than half an hour to observe the change of the operation state of the switch in time, but more than half an hour has passed, The status of the signal light in the switch control panel is still normal. At this time, the author has a kind of intuition that the paralysis of the switch will not happen again. Sure enough, when the author came to the computer room of the training center again the next day, he saw that the signal light of the faulty switch is normal and always in a stable state. So far, the paralysis of the switch has been completely solved
fault summary
although the failure phenomenon of switch paralysis has been eliminated, there are still some doubts in the author's heart. Since the network virus has caused the broadcast storm phenomenon, why another switch with the same model and connection mode does not have paralysis, but only one of the switches has paralysis? Later, when I told another teacher in the training center of my doubts, the teacher reminded me that there must be differences between the two switches, otherwise, there should be no fault in one switch and no fault in the other; After a reminder from the teacher, the author carefully observed the two secondary switches and found that they were different in the degree of old and new. Later, he asked other colleagues to confirm that although the two switches had the same model, one of them was bought before, and the other was bought only when the LAN was reformed recently. According to this difference, the author thinks that e to the relatively long service time of the first switch, its own performance may be aging. Once there is a broadcast storm in the network, it is easy to cause the switch cache overflow phenomenon and paralysis. Because the other switch has just started to use, its own performance is enough to resist the attack of network virus, In this way, the new switch is not prone to paralysis
summarizing the above troubleshooting process, the author thinks that in order to ensure the stable and efficient operation of the LAN, we must select the switch procts with good performance and large cache capacity before networking, so as to improve the anti-interference ability of the switch itself. In addition, in order to avoid the network broadcast storm phenomenon as far as possible, we should install anti-virus software in each computer in the network, and regularly perform the virus removal operation
troubleshooting experience
after a long time and high load operation, the switches in the network will inevitably appear a variety of fault phenomena, so if they encounter fault phenomena without any clue and randomly check, the fault solving efficiency will certainly not be very high. In fact, as long as we have a reasonable troubleshooting sequence, we can effectively improve the efficiency of troubleshooting; Generally speaking, when the switch encounters faults, we can try to check them one by one in the following order:
first, check them in the order of line connection from far to near, because most faults of the switch are often found through the workstations directly connected to it, Therefore, when troubleshooting, we try our best to check in the order of "terminal Workstation - connecting cable - port mole - network jumper - switch"
after checking in the above order, if it is confirmed that there is a fault in the switch, then we have to check the switch equipment in the order from the outside to the inside. We might as well distinguish from the various signal lights in the switch control panel first, and check whether the corresponding components in the switch are faulty according to the fault indication of the signal lights. For example, if the link light of the switch is off, it indicates that the corresponding port is not connected well or there is a problem with the port. If the link light is in the green flashing state, it means that the switch is in the 100MB / s data transmission state. If the link light is in the yellow flashing state, it means that the switch is in the 10MB / s data transmission state at the moment; If the power signal light of the switch is in the green state, it means that the power supply of the switch is normal. If it is in the off state, it means that there is no power supply
when it is confirmed that there is a fault inside the switch, we will not easily dismantle the switch. Therefore, when checking the internal fault, we should start from the system program or parameter configuration. If there is no problem with parameter configuration or system program, it can almost be concluded that there is a problem with the hardware. For example, when a certain port cannot be used normally, we might as well check whether the port connected to the specified workstation is in the corresponding VLAN, or whether the specified port has been closed by other administrators
of course, in the process of troubleshooting the switch, we often encounter some quite complex faults. At this time, we try to start from the system configuration or simple operation according to the order of easy to difficult, and graally analyze and troubleshoot the faults. We believe that this can improve the speed and efficiency of troubleshooting
the steps to quickly find the fault of the key telephone
in order to quickly and effectively find the fault parts of the telephone such as no ringing, no calling and no dialing, the following steps can be generally used for inspection
I. appearance inspection
first, visually inspect the external parts of the telephone, such as the external cable, handle spiral rope and plug-in, whether the fasteners such as the telephone screw are loose, whether there is sound when shaking, whether the spring switch has good press elasticity, ring volume switch, P / T dial transfer switch, anti-theft switch, etc Whether the volume adjustment switch is in the normal position, whether the buttons are stuck, etc< 2. External line voltage measurement: use a multimeter to measure the voltage at both ends of the external junction box of the telephone. For the digital program-controlled exchange, the normal feeding voltage is about 48V when hanging up. If the voltage is lower than 44V, the fork spring switch or ring circuit may be abnormal. When off hook, the normal feed voltage is 6 ~ 10V. If it is too high or too low, it indicates that there is a local open circuit or short circuit in dialing circuit and calling circuit< 3. Check the ringing circuit. The AC voltage of the ring current signal input to the telephone should not be less than 50V. If it is less than 40V, it indicates that the ringing circuit is faulty. It can also use the ring back tone provided by the SPC exchange to test the ringing function. The method is to dial after picking up the phone; 190" Hang up again. If the ringing circuit is normal, a self dialing back tone should be sent< 4. Measuring the working current of the telephone is an essential step in the maintenance of the telephone. Set the multimeter to 50mA DC current, connect it to the telephone line, and measure the working current of the whole machine when it is on hook and off hook respectively
when hanging up, the fork spring switch turns on the ringing circuit. Because there is a DC isolating capacitor at the input end of the ringing circuit, it is normal for the voltmeter to read zero. If there is a reading, it means that the capacitor is leaking. If the off hook current of the telephone is equal to the short circuit current of the subscriber line, it means that the lead wire is short circuited or the fork spring of the telephone is faulty< 5. Check the sending and receiving circuit. Measure the voltage at both ends of the external wiring junction box with the 10 V DC voltage of the multimeter when picking up the phone. Blow air to the transmitter. The multimeter needle should swing normally
if there is a fault in the sending and receiving circuit, check the receiving circuit first and then the sending circuit. When checking the receiving circuit, the dialing tone, busy tone and other signals can be used to find out the fault position through the signal tracking method. For a new type of telephone, after the trouble shooting of receiving telephone, the trouble shooting of sending telephone circuit will also be eliminated< 6. Check the dial and use a multimeter to measure the voltage at both ends of the junction box. Press the number key after off hook. If the pointer of the multimeter swings violently, it indicates that the dialing circuit is normal; Otherwise, the dialing circuit is faulty. If the swing is not big, it means that the dial-up integrated circuit is normal and the pulse circuit is faulty< Check whether the VDD terminal voltage of dial-up IC is 2-5.5v
2. Check whether the startup terminal can be turned over normally (high level on hook, low level off hook)
3. Check whether the oscillator oscillates. Short circuit the oscin and VDD pins. If the oscout terminal becomes low, it means it is normal
4. Check whether there is pulse output at the DP terminal, and measure the pin with the lowest DC voltage of the multimeter. When pressing the number key, if the number of times of the needle swing is the same as that of the number key, it means that it is normal
5. Check whether the switch tube of the telephone switch circuit is normal. The switch tube should be saturated when off hook and cut off when on hook
simple maintenance methods for common faults of telephone set
1. Off hook silent: off hook silent means that the telephone is not working, which is often caused by abnormal DC circuit. The voltage at both ends of external line of local telephone can be measured under off hook condition. The normal value is 6V-10V; If the voltage is very low and becomes 0V, the telephone can be disconnected and the external line terminal voltage of the local telephone can be measured. If the voltage is 48V, it indicates that the input circuit of the telephone has the phenomenon of breakdown and short circuit. The piezoresistor can be checked emphatically, which often has the phenomenon of breakdown in the overvoltage protection< The main reasons are as follows:
(1) the dialing chip can not get normal working current. Normally, after the dialing chip is started. The output voltage of the pulse pin saturates the electronic gate, and the main current of the dial-up chip passes through the voltage stabilizing filter circuit from the electronic gate. If the current limiting resistor or isolation diode of the stabilized voltage filter circuit is open or soldered, the dialing chip will get the normal working current< (2) the clock oscillation circuit does not work. After the Wu Yi vertical input line and Wu Yi horizontal line of the dial-up chip are short circuited, the voltage to ground of the crystal oscillator output end of the element is measured, which is about 1. 5V If it is abnormal, check whether the timing element is in good condition and whether the pin is soldered off. If it is normal, it means that the chip is damaged< (3) the dialing keyboard circuit is abnormal. No pulse dialing, no double frequency dialing
check whether the beginning of T / P conversion is connected well. If there is no voltage at the B pole of the double audio amplifier tube, it can be judged that the amplifier component is faulty soldering or the printed circuit is broken
(4) the handle can't deliver the call. Check whether the handle is broken, and whether the DC bias resistance of the transmitter is increased or soldered
(5) the transmission voice of the handle is small. If the bias voltage of the receiver is about 2V, it is mostly caused by the reverse connection or damp of the receiver lead. At the same time, check the transmitter bias resistance, whether there is deterioration, whether the capacitor is bad, etc
(6) the handle can't accept the call. Touch the B and C poles of the preamplifier tube with tweezers. If there is a "click" sound in the receiver, check whether the spring handle is broken, whether the preamplifier tube and power amplifier tube are in good condition. Whether the output coupling capacitance of power amplifier is bad. If there is a "click" sound in the receiver, check the side sound elimination circuit components< (1)
the phone rings continuously when it is hung up. Generally, the capacitor in the ringing circuit of a telephone is broken down and short circuited, which makes the input of the bell receiver lose its DC function. When the machine is hung up, the feed voltage of the external DC line provides the working power for the ring integrated IC, so the ring will ring continuously when the machine is hung up. In general, just replace the ringing capacitor. If the ringing capacitor is not broken, check the suppression circuit board for leakage or short circuit e to improper solder joint treatment
(2)
the bell rings ring pulse dialing. This is caused by the contact of primary and secondary coils of ringing output transformer. This fault is e to the DC feed current passing through the ring integrated IC after the telephone is off hook. After off hook, the voltage at the outside terminal is low, and the bell receiver will not ring, but when the pulse dialing is on, the amplitude of the pulse voltage is large enough to make the bell receiver ring. Detect the suppression circuit board and solder joint at the output end of ring IC. If there is no collision, replace the transformer
(3)
the ring tone is small. Check whether the DC voltage of the integrated IC is 25 ~ 27V when the bell is closed. If it is much lower than the normal value, check the output coupling capacitor for leakage or breakdown short circuit. If the voltage is basically normal, check whether the output attenuation resistance is increased, and whether the switch and coil are locally short circuited. Otherwise, the IC performance is poor< No ringing
(1)
when any diode in the rectifier bridge is open circuit, the bridge full wave rectifier will become half wave rectifier, which means that the ringing capacitor has only charging circuit but no discharging circuit, so it loses the function of charging and discharging and cannot pass through AC. It can be seen that the ringing current can not pass through the ringing capacitor, so that the ringing IC can not get power and can not work
(2) when there is no ringing fault in the telephone, press the following steps to check in the ringing state
① measure the input AC voltage of rectifier bridge. It is about 60V in normal condition; If it is close to 0V, check whether the ringing capacitor and step-down resistance are open circuit, whether the switch is damaged or whether the lead wire is desoldered< (2) measure the DC voltage of ring IC. 25 ~ 27V in normal condition; If it is close to 0V, check whether the rectifier and filter circuits are broken down and short circuited, and whether there is diode damage in the rectifier bridge, otherwise it is the internal short circuit of ringing IC< (1)
when the telephone rings, it only rings once. After receiving the call, it hears a dial tone and cannot talk. The reason of this kind of fault is that the connection of varistor rv1 is not good or the parameter changes. When the bell rings once, the ring voltage makes the rv1 resistance drop, which is equivalent to the telephone off hook, and the switch automatically cuts off the bell flow. After that, the rv1 resistance graally increases, making the telephone return to the original on hook state. So just ring the bell, pick up the handle can only hear the dial tone. Just change a varistor. In addition, it is possible to prevent the circuit board from moisture, oxidation or leakage. At this time, as long as the circuit board is cleaned and dried
(2) there is a single tone in the telephone ring, that is, there is a continuous "beep -
" sound in the ring, which is the ring distortion fault. This kind of fault is usually caused by abnormal frequency or stop vibration of the ultra-low frequency oscillator. It is necessary to check whether the ultra-low frequency oscillator and external components are in good condition, and whether the ultra-low frequency oscillator has false soldering, short circuit, etc., otherwise, it is the internal damage of the ultra-low frequency oscillator
(3) hoarseness of ring tone is the ring distortion fault, which is generally caused by impure filtering of DC power supply of ultra-low frequency oscillator. It is necessary to detect whether the filter capacitor is invalid or faulty welding, otherwise it is the internal damage of ultra-low frequency oscillator< (1)
when the telephone can only receive the ring, but can not send or receive the call, one of the four diodes in the power directional circuit must be open circuit or short circuit. If the measured DC voltage of the external line is about 48V after off hook, and the voltage is changed to 6 ~ 9V after two external lines are exchanged, there is a diode open circuit in the power directional circuit; If the measured DC voltage of the external line is close to 0V after off hook, and the voltage of the two external lines is adjusted to 6 ~ 9V after switching, there is a diode breakdown and short circuit in the power directional circuit. Just replace the damaged components
(2) * poor contact of spring switch, lead wire desoldering or power supply circuit fault< 5. Pulse dialing is a continuous dialing tone. The disadvantage of pulse dialing is slow dialing speed, waveform distortion and wrong number; The amplitude of pulse signal is large, and it is easy to proce inter line interference. The advantage of al audio dialing mode is that the dialing speed is fast and the signal transmission in carrier telephone system is more convenient. The use of al frequency audio signal can improve the anti-interference ability, rece the connection error of the switch, so as to improve the connection rate of the switch. The al audio dialing mode is especially suitable for SPC exchange
when the pulse dials, you can hear the "click" sound from the pulse and say
2. Do not connect all lower layer switches at one time. It is recommended to use filtering method to connect other lower layer switches
3. Do not connect all computers at one time. It is recommended to connect 1-10 computers at a time and observe the condition of the switch
4. Do not connect any computer or switch, power up the switch directly, connect the switch configuration port, and check whether the switch can save the configuration and start correctly
5. The most common case of this kind of failure is the failure of the lower layer switch he is connected to or the failure of the computer network card
6. Install the firewall to check whether there are LAN attacks
7. I suggest you leave a message for us to study together!
suggestion: specify the IP.. Don't specify DNS and gateway. Because I think what you specify may be wrong
later, the host computer of this computer was carried to another computer that can access the Internet normally. The cat automatic dialing mode and computer dialing mode still can't access the Internet, which shows that the line is absolutely OK
suggestion: this does not mean that there is a problem with the line! Because the circuit and the machine may have problems at the same time, of course, this possibility is not very big. As you said, the protocol, gateway and DNS of the machine are not configured correctly.. Remove TCP / IP and install a new TCP / IP. And do not specify the gateway and DNS. It's OK to specify an IP<
LAN fault
1. Definition example
this kind of fault mainly involves the fault in LAN, broadband network and other network environment
2. Possible fault phenomena
1. The network card doesn't work, and the indicator light status is incorrect
2. The network is not connected, or only a few machines can not access the Internet, can ping but can not connect to the network, and the network transmission speed is slow
3. Data transmission error, network application error or crash, etc
4. The network works normally, but it can't be used in an application
5. You can only see yourself or indivial computers
6
7. Abnormal network device installation
8< Third, it may involve network card, switch (including hub, router, etc.), network cable, motherboard, hard disk, power supply and other related components< 4. Judgment points / sequence
1. Preparation before maintenance
1) available network cable (straight line and ordinary network cable, line sequence in line with international standards)
2) if possible, bring the Internet connection checker
2. Environmental inspection
1) power connection inspection:
A. whether the definition of mains connection is correct
B. whether there is a ground wire
C. whether all devices (such as hub, switch, etc.) on the network have been powered on
2) network cable connection check:
A. whether the network cable connection sequence matches the requirements of network connection (such as direct connection and ordinary network cable)
B. if the connectivity of the network cable is normal, check whether the network cable is damaged or excessively twisted
C. whether the network cable length is too long (for example, the length of type 5 twisted pair exceeds 100 meters required by the technical specifications)
D. whether the cable connector - crystal head is intact and oxidized
E. whether the network card interface is in good condition. Plug in and dial the network cable again to check whether the connection between the network cable and the network card is loose
F. according to the cable requirements, whether there is a terminator and whether the terminator is normal
3) appearance and peripheral inspection of network equipment:
A. whether the network card indicator light is on after power on
B. after the network cable interface of hub and other devices is connected with the terminal or server, if the terminal or server is started and configured normally, its indicator light will be on (pay attention to whether the color of the indicator light is normal, refer to the device manual). If the indicator light is not on, it indicates that the device is faulty
C. whether the network card components are inserted in place without warping, and whether the gold finger on the network card is oxidized
D. whether there is interference around the network cable or switch
4) host appearance inspection:
A. check whether there are foreign matters in the chassis causing short circuit
B. whether there is too much st in the cabinet, and if so, clean the st
C. whether the components on the main board and network card are deformed or discolored
D. after power on, pay attention to whether the components, components and other equipment have peculiar smell, abnormal temperature and other phenomena
5) other aspects:
A. under UNIX, it is necessary to distinguish between terminal crash and server crash
3. Fault judgment points
1) seek the cooperation of user network management
first of all, you should contact the NMS as much as possible to get cooperation from the NMS
2) network environment check:
A. check the compatibility between the network card and the switch for dropped line, packet loss and other faults
B. the network connection is normal, but the domain login cannot be performed. Check from the following points:
A) whether the specified domain name exists or works
b) according to the requirements of the server and the operating system (such as enabling the wins resolution service and DNS service on the server side, WINXP home version cannot log in to the domain.), Set whether the terminal is allowed to log in to the domain and whether the computer name has been registered in the domain
C) check whether the protocol used is correct
C. check whether a firewall is installed and whether it is authorized to access
D. if necessary, use direct connection to connect only two machines, and check whether they can be connected in the peer-to-peer network environment (this can eliminate the influence of various environmental factors on the network)
3) network adapter driver and attribute check:
A. whether the driver is correct and appropriate. It is suggested that the network card device should be automatically identified by the system, and the driver that matches and updates the operating system should be used as much as possible (only the old isa network card can be installed manually). When installing the driver, if necessary, turn off the loaded and running programs before installing them
B. if the network card does not work normally in a certain network environment, the network speed can be adjusted. For example, for a 10 / 100Mbps network card, if the network does not work normally in a 10Mbps network environment, the network card should be specially designated to work at the speed of 10Mbps
C. check the network communication mode, such as whether it is full plex
4) network protocol check:
A. check whether the protocol settings in the network are correct (no matter which protocol is used, the protocol used by the machines in the network must be consistent). Whether there are plicate computer names in the network
B. If you can't see yourself or other computers, first check by pressing F5 to refresh several times, then check whether the file and print sharing service is installed and enabled, and whether the NetBEUI protocol is added (if there is a wins server in the network environment, you don't need to add it, if not, you need to add it)
C. If you can ping the network, but can't access other terminals or servers in the network neighborhood, you can use ipconfig / all (in the command line mode), netstat and other commands to view the specific information, check the settings of network properties, such as domain and working group, and make corresponding changes
D. the utility ping command of TCP / IP protocol can be used to check the operation of the network. This requires the maintenance personnel to understand the relevant knowledge of TCP / IP protocol (Order: Ping 127.0.0.1, local IP, local IP, gateway, DNS, etc.)
E. if Ping doesn't work, you can try to delete all adapters and protocols in the network properties, and then restart and re install them
F. by executing tracert & lt; Destination IP address & gt; Command to check the IP packet in which network segment error
5) system setup and application check:
A. check whether the network card is listed in the list of resources (except non PNP network card) after the self check of the machine, and whether the resources used are shared with other devices
B. check whether there are other devices in the system that conflict with the resources used by the network card. If there are, you can change the installation location between the devices or manually change the conflicting resources. For ISA bus network card, it may be necessary to turn off the PNP attribute of the interrupt in CMOS, and its resources are generally not suitable for sharing with other devices. Older PCI devices are not suitable to share resources with other devices
C. check whether there are viruses in the system
D. if a specific application does not work normally when using the network, check whether the CMOS settings are correct, and focus on whether the driver of the network card matches it. If necessary, close other running applications and start the loaded programs to see whether they can work normally, or compare with the machine that can run the application normally, Check the difference in configuration
E. by re installing the system, check whether the network is not working properly e to the system
6) hardware check:
A. check whether the network card is in good condition with its own program and network card short circuit ring
B. if the network card is still abnormal after replacement, the mainboard can be replaced. If the mainboard cannot be replaced, other network cards can be replaced
7) for diskless stations, pay attention to the following points:
A. whether network boot is allowed in BIOS, it is better to disable floppy drive in BIOS, and set & quot; Report No FDD For Win 95"， Change from yes to no (or vice versa). This is related to the setting of floppy disk drive)
B. for ISA network cards, the BIOS setting should make the boot ROM default starting address D800h or c800h, and I / O 300h (for example, the default setting of some network cards is c800h, which is easy to conflict with the address resources occupied by AGP graphics cards and other components, resulting in installation failure)
C. when the above operations are invalid, some motherboards should be shielded from the on-board sound card, and then corresponding modifications should be made as needed
D. the protocol of workstation must be consistent with that of server
E. when there are multiple servers, the first response server must be specified
8) for wireless network, special attention should be paid to:
A. check whether the effective distance between two terminals is too large, and whether there is an insulator in the middle
B. whether the frequency channels used in peer-to-peer network are consistent
C. in the AP environment, the network ESSID of the terminal must be consistent with that of the AP
D. check whether the keys of the network card and the AP match.
so at the beginning of this period, it's normal, about a month or two, After this probation period, there will be problems
with this in mind, I intend to use other ordinary switches to replace this faulty switch. When I cut off the power supply of the fault switch and prepare to take it out by hand for replacement, I was surprised to find that the surface of the device is very hot, and the temperature of the web server host is also very high. When I put my hand close to these devices
you can also feel the heat flow in the air obviously. It seems that the switch is in a serious "fever" state. Is it the "fever" of the switch that makes other computers unable to access the important host connected to the switch normally? In order to verify my conjecture, I cut off the power supply of the switch and the web server
at the same time, the distance between the switch and the web server host is widened to ensure that they are well ventilated when they are working; In addition, in order to let these "feverish" equipment cool down as soon as possible, the author specially found an electric fan and kept blowing them for half an hour. After feeling that the temperature on the surface of their shell returned to normal, they were re connected to the power supply
at this time, a miracle appeared. The author pinged the IP address of the web server host from his own computer again, and the test results returned this time were all normal; When I open the IE browser and try to access the notification information published by the company on the web server, the browsing speed is very fast. It seems that the switch failure caused by the slow login speed of the server is caused by the serious "fever" of the switch and the web server host
although the switch fault has been successfully eliminated, I wonder why the switch and web server host have such a high fever? After careful investigation of the fault site, the author found that four fans were installed on the top of the cabinet where the switch and web server host were located. Usually, all sides of the cabinet were closed, and only four fans were used for heat dissipation
however, now we don't know why these four fans can't work normally. In this way, the heat emitted by the switch and the web server host can't be removed from the cabinet in time. In addition, the heat emitted by the switch and the web server host will affect each other e to the close contact between them, In the end, their performance is seriously degraded
in order to completely solve the problem of switch failure, the author later invited professional and technical personnel to repair the fan in the cabinet, and at the same time isolated the switch and the web server host, so that the probability of the switch and the web server host "fever" again is greatly reced
although the switch fault has been successfully eliminated, I wonder why the switch and web server host have such a high fever? After careful investigation of the fault site, the author found that four fans were installed on the top of the cabinet where the switch and web server host were located. Usually, all sides of the cabinet were closed, and only four fans were used for heat dissipation
however, now we don't know why these four fans can't work normally. As a result, the large amount of heat emitted by the switch and the web server host can't be removed from the cabinet in time when they work continuously. In addition, the switch and the web server host are in close contact
the heat emitted by them will affect each other, resulting in a serious decline in their performance. In order to completely solve the problem of switch failure, the author later invited professional and technical personnel to repair the fan in the cabinet, and isolated the switch and the web server host. In this way, the probability of the switch and the web server host "fever" again is greatly reced