[mchipser]

Senario:

System works normal then I run a rollup of updates (script to install all updates in a folder), then the network card no longer detects the cable plugged in. We test on 10 models of laptops / desktops and this is the only one with this issue. This rollup is from July of last year, so the KBs / Updates are from that time period.

Model : Compal BlB3 (Transource Roadrunner)
Network Card : Realtek PCIe GBE Family Controller
Operating system: Win7X86 SP1 / Win7x64 SP1 (happens on both OSs)

List of KBs installed in rollup script:
Windows6.1-KB2698365-x86.msu
Windows6.1-KB2655992-x86.msu
vbe62007-kb2596744-fullfile-x86-glb.exe
Windows6.1-KB2691442-x86.msu
msxml4-KB2721691-enu.exe
Windows6.1-KB2719985-x86-6.0.msu
Windows6.1-KB2718523-x86.msu
rvkroots.exe
works6322007-kb2596871-fullfile-x86-glb.exe
infopath2007-kb2596666-fullfile-x86-glb.exe
ipeditor2007-kb2596786-fullfile-x86-glb.exe
msxml4-KB2721691-enu.exe

3rd party items installed in rollup script:
jre-7u5-windows-i586.exe
install_flash_player_11_active_x.exe

Regedits in rollup script:
disable Java phone home: HKLM\SOFTWARE\JavaSoft\Java Update\Policy","EnableJavaUpdate","REG_DWORD","0"

Things I have done to attempt to fix it..
1. Update drivers (no luck)
2. Install all items manually to determine what broke the nic (everything works fine)
3. reset IP stack, I didn't think this would work but i saw it on a forum somewhere

I have attached a copy of the script, this is in autoit. Is there anything else I should try?

Attached File(s)

•  Updates_July12.au3 (9.78K)
  Number of downloads: 0

This post has been edited by mchipser: 18 February 2013 - 08:35 AM

[bphlpt]

mchipser, on 13 February 2013 - 12:11 PM, said:

I have attached a copy of the script, this is in autoit. Is there anything else I should try?

Attached where?

Cheers and Regards

[dencorso]

@mchipser: Do not open two different threads because of the same problem and do not double-post.

[iamtheky]

yeah it probably had nothing to do with the forums being jacked the past week. just as bphlpt couldnt see the attachment, most likely because it still wasnt there 9 hours after being posted, not because he is being dense.

Just saying you could probably give folk a pass on any oddities you notice.

[mchipser]

[bphlpt]

iamtheky, on 14 February 2013 - 09:05 AM, said:

yeah it probably had nothing to do with the forums being jacked the past week. just as bphlpt couldnt see the attachment, most likely because it still wasnt there 9 hours after being posted, not because he is being dense.

Actually, it wasn't there because it was in the duplicate thread the OP had posted, which has since been deleted. Now he has posted the link in this thread as well, so we're all good.

Cheers and Regards

[mchipser]

bphlpt, on 14 February 2013 - 03:12 PM, said:

iamtheky, on 14 February 2013 - 09:05 AM, said:

yeah it probably had nothing to do with the forums being jacked the past week. just as bphlpt couldnt see the attachment, most likely because it still wasnt there 9 hours after being posted, not because he is being dense.

Actually, it wasn't there because it was in the duplicate thread the OP had posted, which has since been deleted. Now he has posted the link in this thread as well, so we're all good.

Cheers and Regards

Actually if you look at the time stamp at which the post was edited, you would see it was before the post asking for the attachment.

This post has been edited by mchipser: 15 February 2013 - 10:45 AM

[jaclaz]

mchipser, on 15 February 2013 - 10:43 AM, said:

Actually if you look at the time stamp at which the post was edited, you would see it was before the post asking for the attachment.

NOT really.
There is must be something buggy in the way the board software manages dates/times.
I already noticed it in some other occasions, the issue is normally because the stupid board software replaces the actual date with "today" or "yesterday", sometimes making absurds things, but evidently there is something else .

RIght now, I see your first post as posted on the 13th, and bhplt's post on "yesterday" (which here means the 14th), but according to your screenshot you posted the first post on 13 February at 12:11 PM.

You then edited it on 13 February at 12:12 PM.

bhplt then managed to reply INSTANTLY on 13 February at 12:11 PM (he COULD NOT POSSIBLY have posted at the shown 9:11 PM quoting something you would have posted for the first time 3 hours later ).

You were cross-posting/cross-editing at the same time, all within one minute between minute 11 and minute 12.

jaclaz

[mchipser]

jaclaz, on 15 February 2013 - 12:24 PM, said:

mchipser, on 15 February 2013 - 10:43 AM, said:

Actually if you look at the time stamp at which the post was edited, you would see it was before the post asking for the attachment.

NOT really.
There is must be something buggy in the way the board software manages dates/times.
I already noticed it in some other occasions, the issue is normally because the stupid board software replaces the actual date with "today" or "yesterday", sometimes making absurds things, but evidently there is something else .

RIght now, I see your first post as posted on the 13th, and bhplt's post on "yesterday" (which here means the 14th), but according to your screenshot you posted the first post on 13 February at 12:11 PM.

You then edited it on 13 February at 12:12 PM.

bhplt then managed to reply INSTANTLY on 13 February at 12:11 PM (he COULD NOT POSSIBLY have posted at the shown 9:11 PM quoting something you would have posted for the first time 3 hours later ).

You were cross-posting/cross-editing at the same time, all within one minute between minute 11 and minute 12.

jaclaz

I believe 12PM is the afternoon which would make his response roughly 9 hours later. Which makes sense, because I posted the original before heading off to lunch on the 13th.

EDIT: This also makes sense, because it is 1:19PM here, and it shows you posted at 12:24PM (afternoon not after midnight).. I am pretty sure the times associated with a post is based on the time zone you are in. This would indicate the time i posted the first message 9 hours later was the response, regardless of the time zone a person is in.

This post has been edited by mchipser: 15 February 2013 - 01:36 PM

[mchipser]

I am not sure how this thread has gotten out of hand, can anyone offer assistance to the problem posted in the original post?

[jaclaz]

mchipser, on 15 February 2013 - 01:17 PM, said:

I believe 12PM is the afternoon which would make his response roughly 9 hours later. Which makes sense, because I posted the original before heading off to lunch on the 13th.

EDIT: This also makes sense, because it is 1:19PM here, and it shows you posted at 12:24PM (afternoon not after midnight).. I am pretty sure the times associated with a post is based on the time zone you are in. This would indicate the time i posted the first message 9 hours later was the response, regardless of the time zone a person is in.

Yep , confirmed.

The real value "inside" your post is:
timestamp='1360779094'

Which should mean:
http://ibboard.co.uk...1360779094&tz=0
13/2/2013, 12:11 with GMT-6
whilst bhplt's post is:
timestamp='1360811499'
http://ibboard.co.uk...1360811499&tz=0
13/2/2013, 21:11 with GMT-6

jaclaz

[mchipser]

Any Ideas why this would happen?

[jaclaz]

mchipser, on 18 February 2013 - 07:55 AM, said:

Any Ideas why this would happen?

UNfortunately, NO.
You seem like having done everything to troubleshoot the issue, the only thing would be (long, and boring, and with no guarantee whatsoever of success) to run the update script after having removed a few updates from it, then restore, remove a few others until you don' t have the effect..

If you already installed all the updates one by one and you did not have that result, it must mean that is the combination of the updates and the script that causes the issue (maybe a timing problem of some kind ).

What you could further do (which I don't seem to see as listed in the attempts) is to uninstall the TCP/IP stack and related other services/protocols/whatever listed in the NIC interface, then uninstall the actual NIC card from device management and reboot.
But it is just a "semi-random" thought/attempt.


jaclaz

[mchipser]

The other odd thing in this scenario is.. If i take a syspreped image of another system which has the updates and apply it to this POS it works fine.

This post has been edited by mchipser: 18 February 2013 - 01:36 PM

[jaclaz]

mchipser, on 18 February 2013 - 01:35 PM, said:

The other odd thing in this scenario is.. If i take a syspreped image of another system which has the updates and apply it to this POS it works fine.

Well that would confirm the guess about a strange timing issue (whatever it might cause it) on the specific system only when running the AutoIT.
IF this timing issue is connected to "software", if you restore to the "queer machine" a sysprepped image coming from another one BEFORE the update, maybe you can re-run the autoit successfully, if it's a hardware originated issue, you should not (unless it was a one-time glitch in the matrix).

jaclaz

[mchipser]

jaclaz, on 18 February 2013 - 02:28 PM, said:

Well that would confirm the guess about a strange timing issue (whatever it might cause it) on the specific system only when running the AutoIT.
IF this timing issue is connected to "software", if you restore to the "queer machine" a sysprepped image coming from another one BEFORE the update, maybe you can re-run the autoit successfully, if it's a hardware originated issue, you should not (unless it was a one-time glitch in the matrix).

jaclaz

The problem is..

We start with image rev A and apply patches to systems that have the image every quarter, until a new rev B comes out. These revs all come from masters or as Microsoft likes to call them the reference machines. I manually update the masters and pull a new image for deployment which include the rollup. This image and the base image all work normal on the machine. I have two of these machines and they are both doing the same exact thing. I have 20 other models and they work normal.

What timing could be the issue? When the patches are applied? I could add a sleep between those, or I can do a runonce and add each one on every reboot

[jaclaz]

mchipser, on 18 February 2013 - 04:05 PM, said:

What timing could be the issue? When the patches are applied? I could add a sleep between those, or I can do a runonce and add each one on every reboot

The most logical (though not necessarily the one that causes the issue) would be some update that "interrogates" the system (either software or hardware) and "spawns" one or more "child process(es)" that are not terminated soon enough, i.e. before the "next" update starts.
The matter must be a question of fractions of seconds (or seconds).
I would try inserting a 5 or 10 seconds "sleep" between each update and see what happens, it will add the most 2 minutes to the update time, which I presume is much more "bearable" than doing several reboots.
Of course actions like removing the java updater should be totally irrelevant.

jaclaz

[mchipser]

jaclaz, on 19 February 2013 - 08:32 AM, said:

The most logical (though not necessarily the one that causes the issue) would be some update that "interrogates" the system (either software or hardware) and "spawns" one or more "child process(es)" that are not terminated soon enough, i.e. before the "next" update starts.
The matter must be a question of fractions of seconds (or seconds).
I would try inserting a 5 or 10 seconds "sleep" between each update and see what happens, it will add the most 2 minutes to the update time, which I presume is much more "bearable" than doing several reboots.
Of course actions like removing the java updater should be totally irrelevant.

jaclaz

It appears this is working.. I am still testing it though..

Thanks!!

[mchipser]

Quote

It appears this is working.. I am still testing it though..

Thanks!!

Well didn't work.. There is another piece to the puzzle..

If the cable is unplugged when doing the rollup, then after the system reboots we plug the cable back in, the cable is never seen to be plugged in. However if the cable stays plugged in while doing the rollup the cable is detected.

This doesn't make any sense.

[jaclaz]

mchipser, on 21 February 2013 - 03:13 PM, said:

Well didn't work.. There is another piece to the puzzle..

If the cable is unplugged when doing the rollup, then after the system reboots we plug the cable back in, the cable is never seen to be plugged in. However if the cable stays plugged in while doing the rollup the cable is detected.

This doesn't make any sense.

I concur.
Really weird.

jaclaz