| Ben さんのプロフィールbenryanauブログつながり | ヘルプ |
|
|
7月8日 Cisco 870 IOS IssuesBeen having troubles with Cisco 870 (877) routers with recent IOS revisions. Earlier revisions had lots of bugs, some nasty and some cosmetic but the real need for a recent release is the improvement in SIP ALG and the addition of several features. Unfortunately the newer releases aren't much better. Seeing serious issues with 12.4(22)T where after a month or so the router runs out of contiguous memory and stops. It doesn't reload - it just becomes incapacitated, which is much worse than a reload. Syslogs show increasing memory fragmentation (rather than a leak per-se). 12.4(24)T is worse - it reloads every half-hour. Haven't been able to grab the console debug yet. Another bloke has blogged similar issues - I've pinged it here http://www.alcatron.net/wp-trackback.php?p=154 Update: c870-advipservicesk9-mz.124-24.T1 is stable, both on NAT and SIP ALG. 3月1日 Defragging MFT, Pagefile, RegistryNice page about defragmentation.. and not just files, but the important stuff. 2月17日 Cisco VPN - WarningA warning if you use Cisco IOS-based PPTP VPN's.
I've got a 2651XM running 12.4(15)T5. I was having terrible trouble establishing a vpn from Visa SP1 - the tunnel would just not work, the "verifying username and password" prompt just timed out. I was using the config I posted in another blog entry which worked fine on a c837.
I tried everything - pulling MPPE and using PAP - to no avail.
Turns out the command "ppp lcp predictive" under the Virtual-Template was killing it - and no debug would clearly show what was going on.
So beware... 2月16日 Free Groupware PM/Task Management Web AppsHaving a need for a small-business groupware solution (specifically for project management, task management and possibly timesheeting, issue tracking) I started to poke around to see what's around. *Redmine
*OpenGoo
*ProjectCompanion
2月15日 MS Terminal Services Client v6.1 Registry LogI'm trying to disable the warning message when connecting to a client - "The identity of the remote computer cannot be verified". Thought I'd record for reference the registry values the TS client (MSTSC v6.1 as with Vista) queries. 2月9日 Windows Mobile 6.1 to Cisco IOS PPTP VPNCAVEAT: ppp lcp predictive doesn't work on all IOS's!
Like the rest of the world, I found Windows Mobile 6.1 would not connect to a Cisco VPDN server over PPTP.
I really needed this, so I struggled with it (for days). Finally I got it working! Sadly this means every customer I need to connect to must have this config.. grrr.
The issue seems to be the WM PPTP client is limited in its capabilities - it isn't very flexible in what it negotiates with the server.
The main trouble was MPPE/MPPC negotiation - WM only supports stateful (or maybe only stateless but not both) MPPE.
PPP negotiation debugs showed a long back and forth CONFREQ CONFNAK sequence where WM just wouldn't accept any offered types.
This behaviour was observed with 12.4(5b) on a c837.
Note you must use MS-CHAP-v2 only on the router. The easiest way is to use MS IAS on the server and configure that properly along with the router's RADIUS config (another blog post later to describe how to get this working!)
In short, here's the relevant Virtual Template config, along with some recommendations for PPTP:
vpdn-group <vpdn group>
accept-dialin
protocol pptp virtual-template 10 pptp tunnel echo 30 ip tos reflect ip tos 15 ip precedence critical ip pmtu ip mtu adjust interface Virtual-Template10
mtu 1380 ip unnumbered Ethernet0 no ip proxy-arp no ip route-cache same-interface ip tcp header-compression peer default ip address pool <your IP pool> keepalive 20 compress mppc ppp lcp predictive ppp encrypt mppe 128 passive ppp authentication ms-chap-v2 <your authentication list to RADIUS> ppp authorization <your authorization list to RADIUS>
ppp ipcp mask 255.255.0.0 ppp ipcp predictive ppp link reorders Let me know how this works for you. And curse you MS and Cisco, so much for the 'pledge to make products work together'.
Keywords: PPTP, VPN, VPDN, LCP, MPPC, MPPE, MS-CHAP-v2, IOS, CISCO, WM, WM6.1, Windows Mobile,, compression, encryption 2月3日 Windows Vista UAC delay (workaround)They joys of consent.exe - User Account Control - stealing your life, one dialog box at a time
A while ago I really started to get sick of the extended delay whenever Vista kicked up a UAC prompt. The delay was between 2 and 10 seconds depending on power management settings (eg throttled CPU on Power Saver profile took longer). So I poked around with ProcMon from Sysinternals to see what was going on.
All I could really see was a heap of registry access related to the audio subsystem, culminating in a "Beep" when the dialog box is finally presented.
So, I stopped the Windows Audio service. Lo and behold - INSTANT UAC PROMPT!
Now if you can explain this, I'm all ears. If you can reproduce this behaviour, I'm even more interested. Needless to say I've been running for a while now with no audio... when I get around to it I'll try Norton's beta UAC replacement tool - http://www.nortonlabs.com/inthelab/uac.php
BTW don't turn off UAC. UAC isn't just a dialog box and a beep - it has a whole shedload of stuff happening in behind the scenes (registry and folder virtualisation, process-sandboxing etc). If you turn it off, you will probably regret it at some point, as well as being less secure.
12月2日 VMware ESX 3.5 Load Balancing IssuesWatch out for issues when using load balancing in ESX. We implemented load balancing / teaming using "Route based on IP Hash" which was tricky enough, but we kept seeing duplicate packets to the guests! Turns out if you set link state monitoring to Beacon Probing you get this issue - set it to Link State (which is less desirable but it resolves the problem). Microsoft Virtual Server 2005 R2 - guest disk issuesEver had a guest in VS2005 show "The device, \Device\Harddisk0, has a bad block." " Windows - Delayed Write Failed" "{Delayed Write Failed} Windows was unable to save all the data for the file" ? I've run into this a few times. Seems the VS host process must have createfile (RWX/FC?) to the host's volume/disk where the VHD's reside (the whole tree up to the VHD files). The account the VM's run under must have access - I create a group "VM Accounts", add my VM user accounts to it and grant FC to eg D:\ (Folder Only), D:\Server (Folder Only), D:\Server\VHD (Folder Only). This resolves the issue. Nice work Microsoft. 9月23日 Vista not appending DNS domain name to lookupsRan across this today..
Vista x32 SP1 box, IE stopped working. It wasn't using the proxy. Trace showed it couldn't auto-detect the proxy. It was attempting to look up "wpad" to grab the proxy wpad.dat file. The local DNS was returning "Server failure". It was because the client wasn't appending the domain suffix to the query. It had been working fine, and no configuration had changed to trigger this issue. DHCP was issuing the suffix with its leases, the adapter had "Append primary and connection-specific DNS suffixes" set.
No "DNS suffix for this connection" was manually entered (this is provided by DHCP).
Nslookup was issuing unqualified names (eg trying to resolve "wpad" not "wpad.local"). All other winsock apps were doing the same.
No funky LSP's were present, nor GPO's or anything in hosts file. The box just stopped resolving, wouldn't append the suffix. I tried disabling/enabling the adapter, flushing the dns resolver cache (ipconfig /flushdns), ipconfig /release ipconfig /renew.
Then after a heap of fiddling it seemed to be better - but I had to close and re-open any apps that were open while the problem was present to get them to work properly. They seemed to have some sort of stateful relationship with (winsock?) that needed to be re-established after whatever I did fixed the global issue.
Weird.. just logging it in case it's a silent issue.
9月12日 looking for thirtythree79 ??..you've found me :) (also listing my username benryanau for google's benefit) 9月3日 Resume from hibernate fails.. - possible hotfix!In the last few days Microsoft have published a request-only hotfix that addresses issues with pcmcia.sys. While the repro case mentioned in the KB article is different to that here, the root cause may be similar in the pcmcia.sys driver. I haven't tested it, mainly due to lack of time to jump through MS hoops to get the hotfix.. (The Hotfix Share may offer an alternative..!) If you do test it, please comment on your results! KB article here - http://support.microsoft.com/kb/957126/ 6月27日 QoS on non-dedicated internet linksThis entry was prompted by a question posed on slashdot. It is meant to shed light on the true issues surrounding QoS in most commercial home-grade links. There's a huge number of people who don't know what the issues with QoS actually are - an unfortunately very common misconception is that configuring CPE QoS features == successful QoS. Also common among the 'enthusiast' group is the belief that sticking in a linux-based QoS-enabled distro is a panacea for QoS (eg "Put in IPCop == solved'").
In a situation where a link is a simple, as-designed ptp situation with no re-encapsulation (eg E1/SHDSL) QoS can be effectively implemented using in-box configuration. Home-class routers and OS-based control will work provided QoS is configured at each endpoint. Unfortunately few people have the kind of service which out-of-box QoS can be used effectively.
The root issue of QoS today is the lack of outbound interface queuing. Unless one can 'see' the hardware/egress packet queue at its final point, QoS will not work successfully (without creating a 'dummy' artificial queue.. see below). When a link is congested, or when smaller frames/packets need to be sent before larger ones, QoS comes into play. First component of QoS is the "queue". This is the 'waiting list' for packet egress (and ingress to a small degree, especially where link-congestion indication is supported). Here is where packets line up to be sent. QoS works by managing the priority of each packet (either by classification or by respecting the IP CoS/DSCP bit in the header). In a simple ptp link, this works well and as intended. Policy is applied to each end, and each router applied a queuing strategy and policy to the packets going out. Bandwidth is known, and the load on the link is directly observable by queue depth. Even applying 'fancy queuing' eg weighted round-robin, Random Early Detection is often enough to ensure equitable access.
Unfortunately in this day and age, most link implementations remove the ability to observe queue depth - by either shaping in the middle, by virtualising the terminating interface, or by multi-hopping the termination.. all three at once is common.
I am discussing DSL specifically as it's the most common access technology, but the same thing applies to cable as rarely can one apply QoS inside the cable CPE.
Originally DSL-based WAN's were intended to be implemented over an ATM-native network, where the CPE configures one or more PVC's and this PVC(s) terminates on the decapsulating aggregation router. The PVC (at each end) has known properties (bitrate, class of service). It was also intended to have multiple PVC's (eg VPI/VCI virtual circuits) with different Class of Service parameters for each to manage QoS. Sadly this model has been discarded, and almost all DSL services 'ignore' the ATM component and just treat it as a 'throwback'. One CPE has one PVC, and this is terminated on an AVGC far outside the reach of the actual ISP with the ATM component having no real impact other than simply being an access transport. To complicate matters, usually shaping is applied at the AVGC end, and the whole path is a series of virtualised interfaces double and triple encapsulating the link. So we have lost the ability to 'see' the actual queues - you can only throw packets at the link and hope they come out the other end.
So now the native ATM opportunity for traffic management is gone, what now? As mentioned, most commercial DSL operates using L2TP multihop and virtual interfaces. A virtual interface has no real queue and in the Cisco world at least, there's no practical method to do anything with the provider end of the link. The QoS mechanism cannot 'see' what the link is doing load-wise and as such can't effectively manage service.
So, what's left?
The only avenue open is to create a 'false queue'. This is effectively an artificial bottleneck. Here, you implement traffic shaping by taking the maximum amount of bandwidth in one direction (and with the bandwidth tax of multiple encapsulations eg PPP, L2TP, ATM which can often be difficult to determine) and subtracting say 10%. This creates a situation where packets can actually queue up and be seen by the QoS process.
It can be done inside some routers, although many home-grade CPE implementations are unacceptable in that they force you to allocate a fixed percentage of bandwidth to a certain service (and often this is TCP/UDP port based which is very limiting in itself).
This is the only way to make QoS work in the current environment. The limitations are obvious: it must be also be done at the provider end, 10% or more loss of bandwidth is incurred, and if there is any congestion between the two endpoints QoS will not be effective.
There is also the option of the CPE router using coarse techniques to 'poison' or control non-sensitive traffic in the presence of time-sensitive traffic. This can be done by using/managing TCP resets, ACK's and window sizes. This approach has many limitations and I cannot recommend it as a practical alternative.
So, if you're after QoS, the CPE router must support traffic-shaping and QoS for upstream control. Your provider must also be willing to provide the same service at the head end. You lose bandwidth, and any congestion between ends can't be accounted for. Pretty horrible isn't it? Until the IT industry gets off its' backside and addresses these issues, the most practical solution for low-latency, low-jitter, low-loss internet is the same as it's always been: get a bigger pipe.
6月18日 Hacking (debricking) the Aironet/Arlan 640 series bridges/AP'sI've a growing collection of the extraordinary Symbol/ARLAN/Telxon/Aironet bridges and AP's.. 630, 640 series. These were pre-802.11 radio AP's and bridges, and are just amazing in their depth of control.
I've a few bricked units I'm trying to revive, and I intend to blog the (slow) progress here.
The main unit is an Arlan 640-2400 2.4ghz. It has a Motorola 68360 series (68EN360FE25C) CPU - 25Mhz Quad Integrated Communications Controller. (datasheet link USER MANUAL) Flash 2 x AM29F010-120JC (1 Megabit (128 K x 8-Bit) CMOS 5v) (DATASHEET LINK) Clock source or EEPROM: Intel 14538B DRAM 2 x 814260-70 RS232 MC145407DW
And some additional glue logic.
Motorola 68360 CPU 25mhz, 4.5 MIPS 8/16/32-bit databus, 32 address lines, glueless SRAM and DRAM interface, IJTAG Access port(!), 7 IRQ's, Four Comms Controllers - Ethernet, HDLC/SDLC 2mbit, UART, two TDM controllers (BRI/PRI ISDN) Parallel/Centronics Interface Instruction Set: CPU32+ (superset of M68000)
Connectivity to EEPROM/FLASH EEPROM (8 bit boot) may be regular or flash. Signals (68k-EEPROM): CSO-CE-Enable, OE-OE(Output Enable), WE0-WE (Write Enable), Data and Address. WE0 - aka UU-WE, Active-low, Address bus A31, corresponds to data bits 31-24
Most interesting is the SIM60 (System Integration Module) as it controls startup, initialisation.
DEBRICKING This unit was bricked by too much debug info being logged - into flash I think. This might have caused a wraparound effect - debug wasn't meant to have full error handling as it's not user-exposed.
The basic plan is to a) try and get a diag on the cause of the boot hang, b) understand the boot process, and c) remedy the cause. I suspect flash has been trashed by the wrap (or it's just plain full) - this would be most easily remedied by unsoldering flash from both bricked and good unit, examining both to see what's there, and if wrap occurred, dumping into a programmer and cloning. Parity might be an issue too. If the issue is EEPROM them same process.
Interfacing to Flash The User's Guide has the interfacing spec on page 9-6, and EEPROM on 9-8.
Boot Process Basic init is described in User's Manual on page 9-10.
<That's all i have the battery for at the moment!>
Collateral to come: M68000PM/AD M68000 Family Programmer's Reference Manual - (link) M68300 Family CPU32 Reference Manual (link)
Update: Unbricking procedures for other routers - short-flash method http://www.dd-wrt.com/wiki/index.php/Recover_from_a_Bad_Flash http://forum.openwrt.org/viewtopic.php?id=1572 http://www.ranvik.net/prosjekter-privat/jtag_for_wrt54g_og_wrt54gs/HairyDairyMaid_WRT54G_v22.pdf - WRT54G EJTAG DeBrick Guide by HairyDairyMaid http://forum.openwrt.org/viewtopic.php?id=5050 - Bit of discussion about JTAG methods http://forum.openwrt.org/viewtopic.php?id=664&p=2 - JTAGging other platforms http://www.dd-wrt.com/phpBB2/viewtopic.php?t=12053 - JTAG pinouts and guides ! ! 5月28日 ATMARP (loopback) Frames being broadcast - Cisco Catalyst 3750 12.2-40 (advipservicesk9)I've noticed one of my switches is broadcasting ATMARP frames every 10 seconds to all switchports.. I had no idea why, as the switch has no ATM interfaces (naturally) nor does the IOS support ATM/LANE. CDP and STP was disabled.
I think these frames are being mis-decoded in Network Monitor - they also showed as loopback frames in the trace (even though classified as ATMARP), which led me to this page http://www.ams-ix.net/technical/config_guide/config_guide.htm#commonly_seen_illegal_traffic_and_setup
Aha! These frames are sent by default every 10 secs on each ethernet interface as a soft connectivity testing mechanism.
This is an example of someone seeing this kind of traffic - http://www.velocityreviews.com/forums/t38199-damn-loop.html
To disable an interface's soft loopback diagnostic frame capability, configure "no keepalives" on the interface.
The downside to configuring this is the interface will always show UP/UP when any media is connected - speed/duplex issues will not trigger UP/DOWN.
Wonderful mysteries and magics of IT :)
5月18日 So Vista doesn't trust you with c:\windows?Vista has set the owner of "system" files to TrustedInstaller and set the ACL to deny anyone from making changes. Cute, but annoying.
Typically anything in c:\windows or c:\windows\system32 will give "Access Denied" even after the myriad UAC prompts.
My Dell Latitude D830 has a 32GB flash-based SSD harddisk, the space constraints of which get a bit tiring.
The performance of the $800 option leave a lot to be desired.. but that's another story. Along with the AHCI issues, BSOD due to LPM changes in Vista.. <sigh>.
Anyhow I wanted to compress the contents of c:\windows\system32\driverstore as it was 1.4GB.. dunno how compressible it will be but we'll see.
Here's a script to fix the permissions issue. Grab it here (plus readme)
Someone at mydigitallife wrote it, it's very basic but automates the changes. Thanks to the author! 5月17日 Resume from hibernate fails with black screen, flashing cursor under Vista- WORKAROUNDHere's a copy of a post I made on the Dell Forums relating to hibernation resume under Vista. Original post - http://www.dellcommunity.com/supportforums/board/message?board.id=vista&message.id=57712#M57712
The issue still exists - I've disabled the pcmcia.sys driver to work around the issue.
---------
Thought I'd post to document what I've found after at least 40 solid hours of troubleshooting.
If your notebook won't fully wake from hibernation, and instead sits there with a black screen with a flashing white cursor at top left of screen, this is probably for you.
The issue specifically occurs here on a Latitude D830 running Vista Ent x32. The command to hibernate completes ok. On power-on, the resume process starts, the "resuming windows" bar winds back and forth with HDD activity, and after 10-20 seconds the video mode changes as expected, but instead of bringing up Windows desktop the machine is stuck/frozen with only a white cursor flashing in a dos-mode display res black screen. It's unresponsive (no kb lights etc) only a long press on power will shut it off. At reboot the NT loader says "Resume failed, do you want to delete restoration data and boot normally".
Unsuccessful troubleshooting included
-All third-party drivers were disabled via regedit (set start type to 4) and all non-essential Microsoft drivers as well (USB etc).
-Firmware and driver versions correct, chkdsk, hiberfil.sys recreated, peripherals removed
-All MS KB articles relating to Vista and hibernation read, relevant hotfixes obtained and applied.. SP1 installed (which was another story)
-All combinations of power management settings
-Replacement of mainboard, processor, RAM, HDD (this was to try and fix another issue with SSD HDD BSOD's, but didn't fix this one)
-Reinstall from scratch several times with Vista Enterprise x32 using only Dell drivers and software
-Left it with the Dell service agents to muddle with (with no success I must add, so I took it back)
Eventually I found what I think is the problem. It seems that disabling the onboard PC card/1394 controller via the BIOS works around this problem. I suspect what is happening is the onboard O2Micro PC Card controller (or the associated smartcard reader and 1394 controller) is buggy in that it does not behave correctly when returning from S4. The supporting driver is the Microsoft pcmcia.sys so it's not a driver issue imo.
This of course is not a solution - it's up to Dell to work with their third-party suppliers to provide hardware that is Vista compliant. I happen to use my PC card slot quite a bit for WWAN and other cards.
Are there any Dell reps here who could recommend what I could do to get this on an engineer's radar?
Hope this helped someone having this problem - if I can I'll update this post with more info as I have it.
Keywords: vista hibernate hibernation resume black flashing cursor hangs freezes "delete restoration data" "Windows Vista Feature Pack for Wireless"Microsoft have announced "Windows Vista Feature Pack for Wireless" - http://support.microsoft.com/kb/942567/en-us
Quote from KB:
"This article describes the Windows Vista Feature Pack for Wireless. This software update includes the following components or features that improve wireless support in Windows Vista:
This update is compatible with all versions of Windows Vista. It can be installed only on a computer that is running Windows Vista Service Pack 1 (SP1)." It is only available (at this time) from hardware vendors/OEM's. Dell seem to be confused by my request for it, they asked me if I wanted to install SP1 (!).
I'm hoping it will resolve my rfcomm.sys BSOD/STOP issues, as rfcomm.sys has been updated in this pack.
BSOD in rfcomm.sysDell Latitude D830, Vista x32 Enterprise w/SP1 and all WU fixes Getting regular BSOD's in rfcomm.sys - including 0x000000D1 IRQL_LESS_OR_NOT_EQUAL Driver Verifier is enabled fully. Bluetooth services installed/enabled include Headset, Activesync and PAN/Internet sharing (via HTC TyTNII/Kaiser)
Microsoft has a March 08 KB article on an rfcomm.sys issue which offers a requested hotfix - this hasn't come through yet so I dunno if it fixes it or not. "When you try to put a Windows Vista-based computer into hibernation (S4), and the computer has multiple USB devices connected to it, the computer stops responding" - http://support.microsoft.com/kb/949537/en-us
There's also the recently announced "Windows Vista Feature Pack" which has a replacement rfcomm.sys in it but this can only obtained from the vendor.. Dell seem to be confused by it but I'll keep persisting with that too.
When I resolve the issue I'll post back.
Keywords rfcomm.sys, STOP, BSOD, Vista, Dell Latitude D830, bluetooth, KB949537
Fix for BSOD Vista Intel driverMy notes on one of the many issues I've had on my Latitude D830 running Vista...
Intel Wireless WiFi Link 4965AGN on Dell Latitude D830, Vista x32 Enterprise w/SP1 and all fixes -No magic firewall or antivirus software (the cure is worse than the disease IMO - using good hygiene I've never had an issue!)
I was getting regular (2-5 per hour) BSOD's when operating on battery, out of dock - mainly when browsing the web (esp IE7.. Yuck i know.). I had enabled Driver Verifier (verifier.exe) fully, and was getting 0x00000019 BAD_POOL_HEADER, 0x000000C4 DRIVER_VERIFIER_DETECTED_VIOLATION, 0x000000C5 DRIVER_CORRUPTED_EXPOOL
I eliminated every non-critical driver (regedit, ccs\services Start=4) including modem, audio, etc but still had the issue. I began to suspect the wireless driver. So I tried another version. Uninstalling and reinstalling wouldn't actually change the driver NETw4v32.sys (2,555,392 bytes - 12/03/2008) After fiddling I did end up getting the machine to BSOD on boot (5-20 secs after logon screen) every time, dunno how. Disabling WLAN in BIOS stopped the BSOD's. This clinched it.
Here's what I did:
-Renamed c:\windows\system32\drivers\NETw4v32.sys to NETw4v32.sysx -Uninstalled the Intel driver package -Double checked NETw4v32.sys was still renamed -Deleted the services via regedit (in HKLM\SYSTEM\CCS\Services) EvtEng, NETw4v32, Regsrvc -Checked no files remained in C:\Program Files\Intel\Wireless -Rebooted with wireless off in BIOS to prevent possible PNP redetection -Checked NETw4v32.sys hadn't returned -Installed Intel PROSet/Wireless driver package v11.5.0.32 (DRIVER ONLY, NOT the IT Administrator package) -Checked file NETw4v32.sys (now 2,251,776 bytes - 26/9/2007) -Reset my personal driver preferences (Adapter properties->Configuration) eg 802.11n=off etc
While this was not the only thing I did to try and resolve these BSOD's, these ones are gone and I am slowly re-adding the other drivers (Bluetooth stack, Virtual PC, WMDC, modem) to ensure it wasn't any of those. I've also reenabled Medium Savings under Power Management.
Interestingly the NETw4v32.sys driver had no digital signature, nor did rightclick properties details show version etc.
Good luck :) |
|
|