- luv-main - example.com

Re: redirecting error on internet access
by zlinw＠mcmedia.com.au 09 Nov '13

09 Nov '13

For the sake of "completeness", I put up the following....... As an experiment I issued the command "pdnsd-ctl empty-cache", this clears pdnsd's cache and resets any local rules. interestingly this cured the problem. It also sped up the internet access a good bit. I believe given what has happened that this would be worth while to do say on a yearly basis. When I first installed pdnsd, I remember the time reduction accessing almost all web pages was very much reduced. I felt over time that this reduction appeared to largly dissapear. I asssumed it did not have as much effect on the 3g broadband, this assumption does not appear to be correct. It seems the longer pdnsd runs the more bogged down it gets, or possibly the more rubbish entries ends up in the cache. It's nice to know my initial diagnosis of a name server problem was correct. Lindsay

3 3

mysqld_safe seems unsafe
by Russell Coker 08 Nov '13

08 Nov '13

It's always a sign of low code quality when a daemon has a wrapper script to restart it when it crashes. Squid used to have such a script but for many years has been considered to be reliable enough not to need it. MySQL seems to be the only commonly used daemon with such a script. I had been wondering if the script was even needed, I would have hoped that mysqld was reliable many years ago. But I checked the logs of a RHEL4 system I run and found 19 log messages about mysqld getting signal 11 since 2006 of which the last one was 4 days ago. One problem with such scripts seems to be the issue of reliably shutting down the server. I have a ZFS system that doesn't always mount the filesystems on boot so I use /etc/rc.local to restart mysqld. That usually works apart from the time when I ended up with two copies of mysqld running at once. However mysqld uses file locking and it seems that you don't get data loss when you have two copies of it running at the same time (try not to test this theory on production networks). While it was handy that mysqld was restarted 19 times, it wasn't good that I was unaware of the problems. Is that script being used in the MySQL successors such as Drizzle? -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/

1 0

LUV's Month of Games
by Lev Lafayette 08 Nov '13

08 Nov '13

On Thu, November 7, 2013 6:44 pm, Lev Lafayette wrote: > > On Thu, November 7, 2013 3:20 pm, Steve Roylance wrote: >> there are some browser games at this site, at least one works in firefox >> on fedora 19. Free to play with some micro-transactions, which are >> mostly avoidable. >> >> http://en.gameforge.com/browser-games > > Great collection Steve, thanks for that. The first post for LUV's "month > of games"! > > I promise to get my talk on the website by tomorrow morning... As promised, MMORPGs and Linux available here: http://levlafayette.com/files/2013-luv-mmorpg.pdf ... and here: https://luv.asn.au/files/2013-luv-mmorpg.pdf Also changed to luv-main, because it is relevant to Linux :) All the best, -- Lev Lafayette, BA (Hons), GCertPM, MBA mobile: 0432 255 208 RFC 1855 Netiquette Guidelines http://www.ietf.org/rfc/rfc1855.txt

1 0

redirecting error on internet access
by Lindsay Sprinter 07 Nov '13

07 Nov '13

I am getting an error on one of my internet access points, if you try and access a web page using say iceweasel one gets a delay of around 3 or 4 seconds then an error message is displayed saying something like "Iceweasel has detected the request being redireted to many times and its not likely to succeed, problem could be cured by enabling or disabling cookies or it could be a server error". Anything else such as an ftp transfer simple gets no response. Now I have a small network here one of the machines acting as the internet access point and firewall. The fire wall is just a small script enacting a "don't speak until spoken to" firewall, the script only acting on a single interface ppp0. There are two access points the first is a 3g mobile broadband modem using Virgin, the second a telephone line dial up modem. This is kept as a backup as the the 3g __CANNOT__ be relied on. Both of these are accessed via the "pon" command, only one is ever used at a time, hence only ppp0 is used. The problem __ONLY__ occurs on the 3g broadband access, the land line being quite OK. Anyone have any ideas on this, it would seem to me that virgin network is doing something strange. Lindsay

1 1

AMD64 graphics driver crash.......
by Lindsay Sprinter 06 Nov '13

06 Nov '13

Hi I am looking for a bit of recent motherboard experience here. I have figured out the cause of the crash's of the Nvidia closed src drivers during my recent i386 V AMD64 experimnents. The problem almpst certainly being a buggy motherboard. The board concerend is a Gigabyte GA-X79-UD5. Its always given a little bit of trouble including problems with the hardrives and one particular graphics card (Radeon HD7870) will not work in text mode, graphics mode though being OK. In both cases the equipment concerned has worked OK on other MB's. Search's on the net also said the some of the early X79 chipsets had some problems. Now I have decided after some thought to replace the MB and have narrowed the search down to 2 examples the first is a Gigabyte GA-X79-UP4, this being an almost perfect direct dropin replacement. The second is an ASUS P9X79 Pro. The reasons for considering the ASUS are firstly I am slightly gun shy of Gigabyte boards due to current problem, second I have two other systems that use ASUS P6X58 pro MB's, these being the previous generation of the X79 boards and these boards have given no trouble and been excellent performers. Third the ASUS boad is slightly more "future proof". Now the question is only minor, the ASUS webs site says the board is only rated for windows 7 and latter although the site does provide drivers for Windows XP. The System has (of course) Linux installed, currently this is Debian Sid i386, with the NVidia closed src driver installed and working OK. Windows XP is also installed, Now the question (finally???) is is there likely to be any problem's installing XP on this newer board. Its not a real great tragedy if one cannot install it but it would be nice to have. I would normaly search the net for the answers for this kind of problem but the dial up line though make this far to slow these days as the net currently expects everyone to have highspeed broadband (my high speed access currently being out of service). Using Linux for now over 20 years, Lindsay

2 1

deactivate vg with offline pv's
by James Harper 06 Nov '13

06 Nov '13

I have a server that has an iscsi connection that should have been disabled, and the volume group is enabled on it. I noticed the volume group after I had disconnected the iscsi connection, and am trying not to add the iscsi back, or to reboot the server. lvscan gave a billion errors but I used dmsetup to manually remove the entries for the lv's. Now I just have the volume group that something seems to know about as when I do vgchange I get: # vgchange -a n vg-drbd /dev/mapper/mpathd: read failed after 0 of 4096 at 3900547923968: Input/output error /dev/mapper/mpathd: read failed after 0 of 4096 at 3900547981312: Input/output error /dev/mapper/mpathd: read failed after 0 of 4096 at 0: Input/output error /dev/mapper/mpathd: read failed after 0 of 4096 at 4096: Input/output error Volume group "vg-drbd" not found I want to tell the kernel to forget that vg-drbd exists altogether as it is no longer reachable and shouldn't be on that machine anyway. Any suggestions? Thanks James

1 1

Postfix log analysis
by Russell Coker 05 Nov '13

05 Nov '13

Does anyone know of a good Postfix log file analysis program to find problems with particular users? In this instance I'm not interested in general statistics or anything. I just want to do searches such as "all attempts for @gmail.com accounts to send mail to user(a)example.com" where example.com is a local domain. The native Postfix logging has one line with a sender address and queue ID and later lines with the queue ID and the recipient(s) which may contain delivery errors. So I can't just grep for the data I want. My problem is when I get complains like "user(a)example.com can't receive mail from friend(a)gmail.com". In many cases it's trivial problems that can be difficult to diagnose, such as friend(a)gmail.com sending mail to user2(a)example.com where user2 has just added the sender to the spam list and not told anyone. So all parts of the server are working correctly but the users aren't getting what they want. -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/

3 3

Re: I386 V AMD64 Debian part 2
by zlinw＠mcmedia.com.au 03 Nov '13

03 Nov '13

Just an update on the current situation.......... It looks like trying to compare i386 to AMD64 on this particular system is a moot point at this stage as I cannot get the NVidia closed src driver to work and the open src driver is __not__ satisfactory for 3D work. It being quite slow and not doing the textures correctly. The problem with the closed src driver is it causes the xserver to exit just after start up due to a memory violation apparently in the pthread library. This happens on all software configurations i have tried, Four different kernels, two versions of Debian (7.2 and current Sid) and 3 differnt versions of the closed src driver. I have posted a bug report onan Nvdia developers forum, there has been no reply as yet. Nvidia docs do say there is somekind of known issue with the pthread library causing crash's so I _MAY_ be suffeering from this. I will have a think about this today, what I may do is to swap the video card. I have a Radeon 2 gigabyte HD7870 spare currently and there are good reports on the latest driver 3.11 beta and this is avaible for Sid. This will save having to reinstall debian. Hmmmmmmmmmmmmmm I may swap the GTX680 into a ASUS X58 MB, this being an older known reliable gamming MB, as the 680 should be OK no reports of problems can be found on the net. Is'nt it a pain when hardware gives problems........... Lindsay

1 0

Wifi performance problem
by Russell Coker 03 Nov '13

03 Nov '13

My home Wifi network used to allow me to FTP files to my Android phones at rates exceeding 1MB/s, I think that 1.9MB/s was my record. With that sort of speed transferring a 200MB or 300MB TV show to a phone was reasonably quick. Performance wasn't always great, sometimes it was as low as 400KB/s, but even that was usable. In the past I had even watched TV shows on my phone over HTTP (from a local server) with no performance problems. In the last few days performance has dropped to below 60KB/s with rates like 40KB/s (that would be 320Kb/s) being common. Because Android doesn't let me know what's happening I connected my Thinkpad T420 via Wifi and saw the following result: # iwconfig wlan0 wlan0 IEEE 802.11abgn ESSID:"abcdef" Mode:Managed Frequency:2.412 GHz Access Point: AA:AA:AA:AA:AA:AA Bit Rate=1 Mb/s Tx-Power=15 dBm Retry long limit:7 RTS thr:off Fragment thr:off Encryption key:off Power Management:on Link Quality=70/70 Signal level=-30 dBm Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0 Tx excessive retries:286 Invalid misc:11 Missed beacon:0 What's going on here? I ran "iwconfig wlan0 rate auto" which made it briefly claim a rate of "130 Mb/s" but then it reported 1Mb/s again. 1Mb/s rate for the Wifi connection would be expected to give something like 40KB/s FTP transfer speed so it seems that the phones are connecting in the same way as my laptop. It says that "abgn" is supported which should give lots of options for rates higher than 1Mb/s, especially when the quality is 70/70 and the Wifi access point (a Netgear DGN 2200 ADSL modem) is about 2M away from my laptop with nothing of note in between. Any suggestions for what I can do to improve performance? -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/

6 6

I386 V AMD64 Debian part 2
by zlinw＠mcmedia.com.au 02 Nov '13

02 Nov '13

During a recent discussion on i386 V AMD64 linux a number of people said stay away from Physical Address Extension PAE kernels, althougth no reasons/evidence was given to back up the warning, What is the problem with PAE. A search on the net showed no major problems ascociated with this extension apart from the obvious limited amount of memory avaible for a single process. As stated I have run PAE kernels now for some time with no problems what so ever. On Debian AMD64 I have managed to get wine running but as yet I cannot compile wine itself, the build complaining of no 32 bit compiler (which is correct). Sadly Debian sid will not let me install one without removing alot of packages including xorg, this certainly not what I want. A search on the net came up with a good bit of info on both running and compiling wine on Debian AMD64 unfortunately none of it was recent and some of the info conflicked. So I am going to have a crack at Debian 7.0.0 i386 which I have a set of DVD's for, having obtained them "just in case" . If its required re intallation of Sid is a non issue with apt-move and saving the package selections. Although there's one minor gripe apt-move will store the AMD64 packages OK but will not handle the i386 items (on the system for wine). The actual amount of data invloved is only around 100meg so its not a real major issue. Lindsay

3 2