Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - Roderick Klein

Pages: 1 ... 27 28 [29] 30 31 ... 44
421

1) various test servers will be impacted by your local tcpip stack settings, for example, take the SACK flag...as I mentioned in my response to Roderick, I did shut it off right now to test it out, the immediate impact => speedtest.net DIES!!! I get a burst of full speed for about 2-3 secs then the traffic completely stops...judging by this result alone I would be inclined to say that having SACK OFF is a mistake...yet hitting several of the other test servers shows no impact, so your results will vary


The advise is based on the first download server that Mensys ever put online for downloads for eCS. I managed 4 servers running OS/2 in a large datacenter with 100 megabit uplink. And with a lot of traffic you could get the TCP/IP stack to get exhaust the system its memory in about 30 minutes by switching SACK on. There is clearly a bug in the stack! Hence the comment on the inetcfg.ini of eCS to *never* turn it on.

Roderick

422
Hi Ian!

First of all, big Thank You for putting me on the right path here. The results so far are great, let me explain step at a time below...

...
Manual MTU value depends on what your fibre is using but you are using mtudiscover for MTU so its fine.

You would normally use 1500 for PPPoA, 1492 for PPPoE but it's really depending on your provider. Some will only accept 1492 for PPPoA and 1452 for PPPoE as your on fibre I would expect you to use 1500 and then let mtudiscover look after it. I have no idea if your provider allows jumbo packets or not, I don't even know if OS/2 supports MTU values above 1500 either but normally smaller is better than larger, up to a point...

I re-set the value to 1500 in my \mptn\bin\setup.cmd. My old ISP which utilized PPPoE over DSL line specifically requested 1452, the new guys...well, I did not get any info because the tech doing the install was very tight lipped..LOL, he specifically asked me to call their tech line instead.

...
If you have created an inetcfg.ini file in your x:\mptn\etc directory.

Note you should change your keepalive and sack values to reflect what is below as well.
...

Why do you suggest that SACK be off? I am basiing this on my understanding of that flag...the following article does a pretty nice job of explaining it's functionality => http://packetlife.net/blog/2010/jun/17/tcp-selective-acknowledgments-sack/. In short, if both the client and server support this, when an actual packet is dropped this will allow the re-transmission of ONLY the dropped data, as opposed to a full payload from the dropped point onwards until the end of the conversation. If successful this implies less data being re-sent.

Regarding the keepalive, I will change it to sync it up with my primary tcpip apps, that being Firefox and Privoxy. Specifically:

1) network.tcp.keepalive.idle_time 600

Beyond the above, what remains is to play with the sizes of the values for these key parameters you list below:

...
The main values you want to play with are:-
tcpswinsize
tcprwinsize
udpswinsize
udprwinsize
...

So far here are the results for the following settings:
tcpswinsize     131072
tcprwinsize     131072
udpswinsize     41600
udprwinsize     41600

My FF based downloads are now moving at a speedy 50 Mb/s, a HUGE improvement from where I started!!!

Various internet connection speed test report a steady 90-95 Mb/s rates, this is substantially better as well.

If you want to see your machine go down quicker, do turn on SACK. We experimented with this setting at Mensys on some servers and as somebody else already reported, stay far clear from switching it on. It will cause more pain then its worth...

Roderick

423
Ian and Remy,

Guys, another update...I thought I would max out the settings with the goal of achieving highest performance...then if I start running into problems I would dial things back a step at a time.

So here are my current applicable INETCFG.INI settings:

#Inetcfg:       CURRENT DEFAULT MINIMUM MAXIMUM

keepalive       300     300     0       7800    KeepAlive (sec)
tcpswinsize     246723  32768   1024    246723  TCP SendWindow Size
tcprwinsize     246723  32768   1024    246723  TCP RecvWindow Size
udpswinsize     246723  41600   1024    246723  UDP SendWindow Size
udprwinsize     246723  41600   1024    246723  UDP RecvWindow Size
lingertime      120     120     0       65535   Linger Time (sec)
ipforward       0       0       0       1       IP Forwarding flag ON/OFF
timestmp        0       0       0       1       TCP TimeStamps ON/OFF
msl             30      30      1       16384   TCP MaxSegLifeTime (sec)
sack            1       0       0       1       Selective ACK ON/OFF

...the xxxwindsize have been set to MAX values.

I still need to research a bit to understand the impact of the following:

lingertime
ipforward
timestamp (I think this may be used by SACK, not sure though)
msl

....and drummroll for the results please!!!  ;D

1) http://speedsmart.net/ => 87 Mb/s
2) http://testmy.net => 40 Mb/s
3) http://beta.speedtest.net => 91 Mb/s
4) FF download => 5000 KB/s, so roughly sustained 50 Mb/s

All in all, I would say that is basically what I'm seeing on my Winx boxes.

It makes sense... Windows XP also had IP stack tuners to change some of these value's as you had reduced performance duo to certain IP settings not being tuned automatically.

Roderick

424
Networking / Re: Intel 2915, no networks
« on: June 21, 2018, 01:24:41 am »
So the best of my understand the XWLAN in eCS 2.1 is antique so an upgrade is the first good step.
The issue you describe is a known bug. There are to many WiFi access, that is why your list is empty. This was fixed in later versions of XWLAN. 

One important thing is that the Xlwan should be able tind the file genmac.dim, which is bascially a DLL to access the Genmac driver. That is the only thing I can think off why it can not find the hardware.

Did you also check lantran.log that the driver really loaded succesfully ?

425
Networking / Re: NetDrive traps - since latest LIBCX drop...
« on: May 30, 2018, 04:40:14 pm »
Since the latest libcx/libc release (0.6.2-1.oc00 / 0.6.6-36.oc00) I started to see a consistent NetDrive crashes:

Code: [Select]
------------------------------------------------------------
04-28-2018  18:44:32  SYS3175  PID 0021  TID 0003  Slot 0061
G:\UTIL\NDFS\NDCTL.EXE
c0000005
1e754d45
P1=00000001  P2=2105002c  P3=XXXXXXXX  P4=XXXXXXXX 
EAX=0289f92c  EBX=006f5380  ECX=00000001  EDX=00000004
ESI=2105002c  EDI=0289f92c 
DS=0053  DSACC=f0f3  DSLIM=ffffffff 
ES=0053  ESACC=f0f3  ESLIM=ffffffff 
FS=150b  FSACC=00f3  FSLIM=00000030
GS=0000  GSACC=****  GSLIM=********
CS:EIP=005b:1e754d45  CSACC=f0df  CSLIM=ffffffff
SS:ESP=0053:0289f8ac  SSACC=f0f3  SSLIM=ffffffff
EBP=0289f8d0  FLG=00010212

LIBC066.DLL 0001:00054d45
------------------------------------------------------------

Has anyone else noticed anything similar to this?

I wonder whether the newly introduced 'SET LIBCX_HIGHMEM' option is causing some problems?

Currently the whole OS install runs with the default, which I believe is to do memhigh allocation whenever possible. I have not tried defaulting to 'SET LIBCX_HIGHMEM=2' instead, although that is my next test to see if NetDrive is impacted in any way.

Thanks!

Did you open a ticket with BWW about this ?

Roderick Klein

426
Networking / Re: Wifi bridge for OS/2 machines
« on: May 30, 2018, 04:34:44 pm »
I am using a TP-LINK TL-WR802N here in a hotel to to access the wifi.
Look up the model number on google.com

Roderick Klein

427
Networking / Re: Intel 2915ABG on a T41 with OS/2 (questions)
« on: May 30, 2018, 04:32:52 pm »
So I finally got my T41 upgraded and thrown back together. The Intel 2915ABG wifi card is installed, and the BIOS has been modded to disable the whitelisting. IBM states in their documentation that the 2915ABG is not supported on the T41, only the T42 and beyond (however I presume that's due to the said whitelisting). The BIOS did complain that an unsupported Wifi module was detected prior to modding the BIOS.

Whenever I boot into OS/2 (ArcaOS) I only see the wired Intel NIC, when I disabled the wired Intel NIC OS/2 then complains because it's no longer present and the way it configures things is very stepped like that.

So... my question is...
Should I reinstall ArcaOS if it needs to see network devices during the installation process to set up? And if I do, how exactly does one get GenMAC going?

I'm presuming that the T41 is in fact compatible with the 2915ABG.


I think it should be able to install the Wifi driver next to the wired NIC (and auto detect it). If not open a ticket at https://mantis.arcanoae.com (you can login with your userid and password).
Also what version of ArcaOS have you got 5.02, 5.01 or 5.00 ?
(You can tell by checking your ArcaOS  DVD the file \SYS\arcanoae.flg)

When you open a ticket follow the following instructions:
https://www.arcanoae.com/wiki/reporting-problems-best-practices/
Possibly the database does not have the entry to detect the NIC.

Best regards,

Roderick Klein
President OS/2 VOICE

428
The way I am reading your it quickly you used the throttle option. Be very careful with this as is written in the manual of the PSD!

WARNING: THROTTLING IS NOT RECOMMENDED The PSD has other mechanisms for reducing power consumption and reducing temperature. Throttling the CPU in combination with these other mechanisms can actually increase power consumption and increase the processor temperature. Throttling the CPU can also cause other problems such as reboot failures, power off failures, etc. If you are using CPU throttling and have any (even seemingly unrelated) problems, disable throttling. Remember, THROTTLING IS NOT RECOMMENDED and is known to cause problems. Use it at your own risk.

What is the make and model of this system BTW ?

429
Hi Roderick and thanks for the reply!

It's frustrating because they change the Linux kernel driver API too often, sure. The latest news have been that BSD variants are most likely to die out in a long run and Linux will dominate the open-source Unix clone market.

I do not know if FreeBSD is going to do die out there was a whole a lot of noise about I just found via google.
But I could be wrong....

To answer your question about a wrapper driver to load Windows Wifi drivers. We have that on OS/2 in the form of Genmac. But the development stopped as the Windows NIC drivers started to change to much.

Roderick

430

<SNIP>
Hi All!

So, I decided to buy an older Thinkpad as my ArcaOS machine. I sold my old T400 a couple of years ago when I switched to Apple laptops so I needed to get a new one. The first Thinkpad my vendor had to offer was a T410s that wasn't a complete success because the WLAN wasn't supported and there was an issue with graphics: windowed DOS or Win/OS2 sessions didn't work even with the suggested CONFIG.SYS and AUTOEXEC.BAT modifications. So, I decided to sell the T410s. I really liked the specs and appearance of the machine BUT getting ArcaOS to run perfectly was the goal. Next I got a bigger pile of older and cheaper machines: T60, T61, T400, X200 and X201. According to Neil Waldhauer the T60 would be a perfect ArcaOS machine and X200 would also work but the WLAN card has to be changed a specific Intel one. I've tested T60 and X200 and the information seems to be correct. On T60 everything works out-of-the- box and with X200 everything works except for the WLAN. I also wanted to have a fast ArcaOS desktop (I have two licenses) so I bought a used Lenovo all-in-one: ThinkCenter M90z, however there are issues with the integrated graphics.


WLAN is a known issue for OS/2.  When I still worked at Mensys I hired a Russian company to port the E1000 gigabit driver to OS/2 from Linux (Please note that I left Mensys in 2013 I am no longer associated with eCS in any way). Later this project moved on using a FreeBSD based driver library (I picked up this knowledge from David Azarewicz his Warpstock presentations.  All of the newly ported NIC drivers are compiled FreeBSD kernel sources. The issue why wireless is taking so long is because the Linux port of the NIC drivers at Mensys at the time was cancelled duo to the fact that some things in the Linux sources changest to radically that David started basically from scratch with the FreeBSD code. The NIC driver I think are working pretty good for ethernet. But wifi is extremely more complicated that ethernet (wired chips are).

Other drivers seem to be doing pretty well. Arca Noae is working on a UNIAUD replacement audio driver and USB 3.0 seems to have progressed extremely far.

I really do not understand why you did go for example with a Thinkpad T540p.
http://www.blondeguy.com/computer/strobopeep.html

As Arca Noae said many times if you have an issue with installing ArcaOS open a ticket in https://mantis.arcanoae.com.
Arca Noae is in the same position as I was at the time at Mensys. They do not have access to all hardware.
That said computer hardware has become more generic making it somewhat easier to support hardware. Its still a lot of work to develop the drivers. But once a driver works a lot of system should work.
If you have an issue open a ticket and follow these instructions:
https://www.arcanoae.com/wiki/reporting-problems-best-practices/

Roderick Klein
President OS/2 VOICE

431
Hi All!

<SNIP>
Unfortunately the Firefox port seems to be garbage. It's quite unstable and SLOW. When the system itself runs blazingly fast, loading the browser takes a long time. The 38 version has a nasty bug when viewing Youtube videos full-screen and supports only 360p resolution. The 45 version doesn't like to play videos at all (it plays for 1-2 seconds and then stops), the full range of resolutions is available there but it doesn't help much because the videos refuse to load. The future of web browsing in ArcaOS is a bit uncertain... What would be the best browser in the future? I don't have very high hopes with the Pale Moon browser because of the bad experiences with current Firefox builds which it's most likely based on. Maybe something Webkit based would be better in the future? Some more in-depth documentation is also needed. For example, I didn't find any way to switch the graphics driver from SNAP to Panorama or vice versa and the only solution I came up with was to reinstall whole ArcaOS. Printing works with CUPS but not from 16-bit Windows applications. The sound in WinOS/2 doesn't work either. I tried an UNIAUD WinOS/2 driver but it does't work. Sometimes the "tada.wav" can be heard when individual 16-bit Windows program is started but that's all.

But still I like using ArcaOS very much and I'm looking forward for future versions. I'm sure that the current issues will be fixed in time and I like to do what I can to support the effort. Or am I the only one having the issues mentioned? Or might it be the case that I've missed something in documentation or in forum posts?

When it comes to the browser discussion I suggested that you read what I just posted in this thread about the browser discussion:
https://www.os2world.com/forum/index.php/topic,1726.0.html

Based on what I heard from other users Firefox 45 does seem to be getting more stable.

Roderick Klein
President OS/2 VOICE

432
I just like emphasize again that this funding campaign was started October last year. Its true it was not clear 100% at he time which  direction it would go. The QT based browser was the most likely candidate. VOICE started collecting funding for BWW early as it was expected this would be a big effort to get this project moving.
And why should we always be late at the party as a community!

Some people where not happy with this direction. Dmitry from Bitiwise works looked in some of the options and QT based browser seems to be the best approach. Do not forgot that the guys at working on porting code AND have to do this research in parallel what direction to take. This is the reason it took time to make the QT roadmap.
It again extremely important to understand that which browser we choose is not the primary point of discussion at this time as 85% to 90% a QT based browsers is "tadaa" coming out of the the QT libraries.

Firefox has had its best days. The following email is quoted from Dmitry from bww bitwiseworks GmbH:

"Many years of developing Firefox for OS/2 (since version 10 when it was officially dropped by everybody and bww bitwiseworks GmbH picked it up, till the latest version 49) proved that this project is too overwhelming for a small group of OS/2 developers that we have nowadays. The Firefox source code is a huge mess of everything — third party libraries, new components developed from scratch by various groups of people inside the Mozilla corporation, old parts coming from Netscape times. Firefox contains a dozen thousand source files. The Mozilla corporation has 1000+ employees working on the components of the Mozilla infrastructure that Firefox is made of. It's simply not possible any more to bear its complexity with our small developer's community. Its not possible to do this in a reasonable amount of time and stay in line with the Mozilla mainstream development progress.

For this reason it was decided to stop spending time and money on porting a project whose quality cannot be guaranteed on OS/2 any more. More over, there are alternatives. The most interesting one is Qt software development framework version 5 that among other things contains the web browser sub-module based on the Google's Chromium engine (the same one used in the popular Chrome browser). Having Qt 5 and its web browser sub-module ported to OS/2 will give us not only a number of Qt-based web-browsers but also a whole bunch of Qt-based applications for various purposes. We have a port of Qt 4 to OS/2 but it's became very old and most Qt applications moved to Qt 5 long ago. Also having the Chromium ported to OS/2 will open doors to having a port of Chrome to OS/2 if the need arises."

Anyway that is summary from Dmitry and I think he is the best guy we have for the job to know what he is talking about. He has been busy porting Firefox for years and he knows a lot about OS/2.


Roderick Klein
President OS/2 VOICE

433
Setup & Installation / Re: OS/2 Warp 3 Installation Issue
« on: March 23, 2018, 11:08:12 pm »
When you get the OS/2 boot blob press ALT F2 key combination to see which driver its loading.
The fact that DOS boots is not a garentee that OS/2 will boot since DOS leans on the BIOS to run.
OS/2 uses its own drivers. Warp 3 is from 1994 I do not know if without updating the boot diskettes with new IDE driver etc. You can get this to boot. Warp 3 is to long ago for me, sorry.

434
Setup & Installation / Re: SYS 1201 error during boot.
« on: March 23, 2018, 07:50:19 pm »
Sometimes I see driver boot failures when memory is exhausted. I am not talking about the normal memory pool the end-users see though, so maybe this has something to do with shared memory?

Here is what I see (as an example): all my partitions are HPFS386 (there are 15 of them), when I am trying a new device driver, often times the HPFS386 cache size (set to 128Meg right now) becomes too large and during boot-up I start seeing HPFS386 disk error messages...well, no errors exist, but during the boot process, whatever changes I implemented caused the available memory for HPFS386 cache to be exhausted, thus the failures.

Another example of this is playing with SNAP allocated video memory amount. The hardware has 256Meg on-board, but I only have it set to use 32Meg, else I run into problems booting various other device drivers.

Sooo...my point being: could what you are seeing be similarly driven by memory availability? Maybe the new AOS drivers are now consuming more?

If you boot your OS/2 system with many of the larger driver such as ACPI.PSD, Uniaud, SNAP and a few JFS drivers.
Always, always set the setting earlymeminit=true. If do you do not you have 16 MB of RAM to load all drivers.
Note that this switch does not work that well with HPFS386.IFS.
The issue is that when the system runs out memory it can simply crash during startup somewhere.

Roderick

435
The only reason for explaining this in more detail is I sometimes see being mentioned in forums "disable the disc checker" as it does not help people to trace the true cause of the error. Or fix the issue.

Here are some scenario's where the disc checker does work.
If you have a hard disc where one partition is created without LVM/minilvm it will not have LVM DLAT information.
Even if a partition does not have a drive letter assigned it will require a DLAT information. If this is not present LVM will back fire with a partition corrupt message. If you do not have a tool such as dfsee you are stuck.

If you have a hard disc between 512 GB and 2 TB in size and have an OS installed such as Windows it will guide you to make space in the proper way.

It can also detect if a partition is outside the boundaries of what OS/2 can handle in terms of disc geometry.
Is it perfect ? No its... I know of cases where it does detect certain disc layout errors...

Anyway that is what I remember of when Alex Taylor developed that tool. But that is some years ago...

Roderick

Pages: 1 ... 27 28 [29] 30 31 ... 44