Re: [gentoo-user] KDE 4.10 and plasma crashing
Dale rdalek1...@gmail.com wrote: Frank Steinmetzger wrote: Am 08.02.2013 23:54, schrieb Neil Bothwick: On Fri, 08 Feb 2013 16:45:13 -0600, Dale wrote: Well, switched to a newer gcc, same thing. Going back to KDE 4.9 for a bit. They will have it fixed in a couple days. After all, Linux has some of the smartest programmers there is. I'm not sure of some users tho, myself included. ;-) Changing CFLAGS and rebuilding one package is a lot less work that a complete downgrade. Only if you don't have a backup (which I did before upgrading, but I can't be bothered with redoing the upgrade in only a few days, so I'll sit it out with Awesome in the meantime). Thankfully, my netbook, which needs very many an hour to build KDE, runs x86 and isn't affected. Yep. I rm'd the kde.keyword file and did a emerge -kuv world and down went KDE. Took about 5 or 10 minutes. I cooked and ate supper while the drive light blinked. Back to normal now. Thanks again. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words! Of course I have backups, although a quick run of demerge got it back for me. But rolling back KDE, with the attendant hassles of messed up configs, only sidesteps the bug , which relates to qr-core not KDE. At least it prompted me to, belatedley, set up snapshots on my new ZFS setup, so some good came of it. -- Sent from my Android phone with K-9 Mail. Please excuse my brevity.
Re: [gentoo-user] neon/davfs2/cadaver failing with openssl-1.0.1c
Have you tried revdep-rebuild -L libssl.so.1.0.0 ? - Mail original - De: William Kenworthy bi...@iinet.net.au À: gentoo-user@lists.gentoo.org Envoyé: Samedi 9 Février 2013 01:33:42 Objet: [gentoo-user] neon/davfs2/cadaver failing with openssl-1.0.1c I have been using dev-libs/openssl-0.9.8x with neon/davfs2 to mount a share from blackboard, but after upgrading to 1.01c it no longer works. Ive been rebuilding and cleaning up the libs but no luck so far. The problem is limited (as far as I can see) to neon and the autofallback to tlsv1.0 failing (server will only do that version). openssl-1.0.1c also installs libraries numbered like /usr/lib/libssl.so.1.0.0 - its probably not related though one of two failing systems did have a 1.0.0 version installed at one stage. Has this been seen before? BillK
Re: [gentoo-user] Re: multiple installs
You could install to one of the Athlons, then copy it to the other two Athlons and one of the FX machines. Then, reconfigure the FX install for more CPUs and native -march, recompile world and copy it to the other two FX machines. Even that might not be worth it. Unless you're using something that benefits from the new instructions in the FXes, or requires the last .1% performance, a single install on an Athlon and copy onto the remaining 5 systems will give you more time for something else. From a support point of view the consistency can be a win. Just remember to specify the Althlon64 CPU type in the kernel config, but with 8 cores max. For the hardware, just select a superset of the two systems requirements in your kernel config. If you build as modules any superfluous code wont even be loaded.
Re: [gentoo-user] {OT} LWP::UserAgent slows website
There are several things you can do to improve the state of things. The first and foremost is to add caching in front of the server, using an accelerator proxy. (i.e. squid running in accelerator mode.) In this way, you have a program which receives the user's request, checks to see if it's a request that it already has a response for, checks whether that response is still valid, and then checks to see whether or not it's permitted to respond on the server's behalf...almost entirely without bothering the main web server. This process is far, far, far faster than having the request hit the serving application's main code. I was under the impression that Apache coded sensibly enough to handle incoming requests as least as well as Squid would. Agree with everything else tho. OP should look into what's required on the back end to process those 6 requests, as it superficially appears that a very small number of requests is generating a huge amount of work, and that means the site would be easy to DoS.
Re: [gentoo-user] neon/davfs2/cadaver failing with openssl-1.0.1c
Many times! (and the list includes libreoffice so its painfully long in build time :( I had one machine working still using 9.8x so I upgraded that ... and it broke too. ok, so I thought, go back to 9.8 and remove 1.0.1c ... and found that its now slotted, and is just a stub so it installs libs but no headers etc ... useless when you want to downgrade :( There is an entry in the changelog saying 1.0.1c suffers breakage (but not what) so I went ~x86 and its rebuilding now, but still doesn't work. Cant believe I am the only one suffering this, unless its just some servers like the blackboard ones :( BillK On 09/02/13 17:47, Arnaud Desgranges wrote: Have you tried revdep-rebuild -L libssl.so.1.0.0 ? - Mail original - De: William Kenworthy bi...@iinet.net.au À: gentoo-user@lists.gentoo.org Envoyé: Samedi 9 Février 2013 01:33:42 Objet: [gentoo-user] neon/davfs2/cadaver failing with openssl-1.0.1c I have been using dev-libs/openssl-0.9.8x with neon/davfs2 to mount a share from blackboard, but after upgrading to 1.01c it no longer works. Ive been rebuilding and cleaning up the libs but no luck so far. The problem is limited (as far as I can see) to neon and the autofallback to tlsv1.0 failing (server will only do that version). openssl-1.0.1c also installs libraries numbered like /usr/lib/libssl.so.1.0.0 - its probably not related though one of two failing systems did have a 1.0.0 version installed at one stage. Has this been seen before? BillK
Re: [gentoo-user] KDE 4.10 and plasma crashing
On Sat, Feb 9, 2013 at 9:58 AM, Nilesh Govindrajan m...@nileshgr.com wrote: Dale, which version of GCC are you using? I'm on 4.7.2, updating KDE now, will report back after a few hours when it's done. -- Nilesh Govindarajan http://nileshgr.com Those few hours took almost the whole day, meh! Thanks to chromium. Anyways, it's working without any hassles for me. -- Nilesh Govindrajan http://nileshgr.com
[gentoo-user] Getting back new config files
Hi, Quite a lot of times there are config updates to be performed using dispatch-conf. Suppose I discard a config by mistake while updating one but I know which package it belongs to. How do I get the new configuration back? Remerging the package doesn't seem to solve the problem. I'm using paludis. -- Nilesh Govindrajan http://nileshgr.com
Re: [gentoo-user] Getting back new config files
On Sat, 9 Feb 2013 18:13:18 +0530, Nilesh Govindrajan wrote: Quite a lot of times there are config updates to be performed using dispatch-conf. Suppose I discard a config by mistake while updating one but I know which package it belongs to. How do I get the new configuration back? Remerging the package doesn't seem to solve the problem. I'm using paludis. emerge has the --noconfmem to deal with this. I don't use paludis but expect it has something similar. -- Neil Bothwick This message has been cruelly tested on sweet little furry animals. signature.asc Description: PGP signature
Re: [gentoo-user] {OT} LWP::UserAgent slows website
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 02/09/2013 05:36 AM, Adam Carter wrote: There are several things you can do to improve the state of things. The first and foremost is to add caching in front of the server, using an accelerator proxy. (i.e. squid running in accelerator mode.) In this way, you have a program which receives the user's request, checks to see if it's a request that it already has a response for, checks whether that response is still valid, and then checks to see whether or not it's permitted to respond on the server's behalf...almost entirely without bothering the main web server. This process is far, far, far faster than having the request hit the serving application's main code. I was under the impression that Apache coded sensibly enough to handle incoming requests as least as well as Squid would. Agree with everything else tho. Sure, so long as Apache doesn't have any additional modules loaded. If it's got something like mod_php loaded (extraordinarily common), mod_perl or mod_python (less common, now) then the init time of mod_php gets added to the init time for every request handler. OP should look into what's required on the back end to process those 6 requests, as it superficially appears that a very small number of requests is generating a huge amount of work, and that means the site would be easy to DoS. Absolutely, hence the steps I outlined to reduce or optimize backend processing. -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.19 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQEcBAEBAgAGBQJRFlRSAAoJED5TcEBdxYwQ7BwH/Aj3hgQgGjzBoQhlZqPKDzEW pZJJVcVf4CF4sk88el8X/hPMfx2cTpuM53tLDsv3KGR1dwjP48O2oiiTubH/HRxI lNR5I22QK2YEbLzeRTZN+pkpGnyA1W+d3kF7F9aiNXVUV8KyuyxSxx+7Xm1tRW/W xcNhSLTQIpyTAx+R9MGNkJFs0gFGFgIMML4bfi5BpIrbeeVWsoe1C0syFF+HIFWP WZRtsCFhdWrZkvKUYIBkoFq9VKkSTt13eIvrPjxFUVJwFSmntxSgfqiaZxfHXp5A oSLtyz0vR6qByoivkuilNK7sI3fK8fHA0q4XF1AUaOuwcHg9AFG9pCFBUF2KOgk= =R/kD -END PGP SIGNATURE-
Re: [gentoo-user] KDE 4.10 and plasma crashing
Neil Bothwick wrote: Dale rdalek1...@gmail.com wrote: Frank Steinmetzger wrote: Am 08.02.2013 23:54, schrieb Neil Bothwick: On Fri, 08 Feb 2013 16:45:13 -0600, Dale wrote: Well, switched to a newer gcc, same thing. Going back to KDE 4.9 for a bit. They will have it fixed in a couple days. After all, Linux has some of the smartest programmers there is. I'm not sure of some users tho, myself included. ;-) Changing CFLAGS and rebuilding one package is a lot less work that a complete downgrade. Only if you don't have a backup (which I did before upgrading, but I can't be bothered with redoing the upgrade in only a few days, so I'll sit it out with Awesome in the meantime). Thankfully, my netbook, which needs very many an hour to build KDE, runs x86 and isn't affected. Yep. I rm'd the kde.keyword file and did a emerge -kuv world and down went KDE. Took about 5 or 10 minutes. I cooked and ate supper while the drive light blinked. Back to normal now. Thanks again. Dale :-) :-) Of course I have backups, although a quick run of demerge got it back for me. But rolling back KDE, with the attendant hassles of messed up configs, only sidesteps the bug , which relates to qr-core not KDE. At least it prompted me to, belatedley, set up snapshots on my new ZFS setup, so some good came of it. -- Sent from my Android phone with K-9 Mail. Please excuse my brevity. I didn't have any config problems at all. I did have to update but its the same thing I have to do when I do a upgrade. It may only be a qt problem but it sure did bork KDE up badly. KDE was practically dead in the water. It reminded me of a ship that got hit by a large torpedo. It was not sinking, it was sunk. Almost nothing KDE worked. No kicker, no wallpaper for the desktop, no KDE menu, no right clicking on the desktop. Also, no way to log out either. The only way to logout was to restart xdm. Yea, no matter what caused it, for ME at least, downgrading was the easiest and fastest. YMMV tho. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] Re: Bug in spidermonkey?
Hi Walt On 2013-02-08, walt wrote: I just built 1.8.5-r1 on my ~amd64 machine without errors, and I have 1.8.5-r4 already installed. Can you unmask and compile r4 just as a test? I tried the r4 version but it failed the same way. BTW, in your build log I see waiting for unfinished jobs, so it would be worth trying -j1 instead of -j4. Your stable build tools are no doubt very different from the versions I'm running on ~amd64. Just another random thing to try in a puzzling situation... I also (additionally) set the MAKEOPTS to -j1 and it failed as well. I guess I'm going to reinstall the whole system. -- Greetings Elias
[gentoo-user] Re: multiple installs
On 02/08/2013 11:46 AM, James wrote: Hello, I have quit a few amd64 systems to install, hopefully mostly unattended. I'm looking for a way to install quick and simple workstations running kde. All will have (boot, root and swap partitions only). They can be updated to current, individually. later. http://en.gentoo-wiki.com/wiki/Install_LiveDVD_11.2_to_hard_disk_drive So the best guide I have found is this one for the livedvd 11.2. Any modifications for the livedvd-amd64-multilib-20121221.iso version? Grub-2, udev, and some other changes may make following this 11.2 guide, problematic? I wanted to do these installs with ZFS The only obvious problem I can see is that grub2 will need zfs support if your /boot is going to be zfs. I don't recall all of the details, but at one point during the grub2 install you can tell it to pre-load the zfs module (and any other modules you may want) during the install.
Re: [gentoo-user] Which CPU type to select?
On Fri, Feb 08, 2013 at 08:48:08PM -0500, Walter Dnes wrote: I'm installing Gentoo on a brand new toy I just got myself. Here are the choices from make menuconfig. Is Core 2/newer Xeon correct? ( ) Opteron/Athlon64/Hammer/K8 ( ) Intel P4 / older Netburst based Xeon (X) Core 2/newer Xeon ( ) Intel Atom ( ) Generic-x86-64 Here's the listing for one of the cores from /proc/cpuinfo vendor_id : GenuineIntel cpu family : 6 model : 37 model name : Intel(R) Core(TM) i7 CPU M 620 @ 2.67GHz You can also use this patch which adds support for modern (=gcc-4.6) CPU options to the kernel: http://www.linuxforge.net/linux/kernel/kernel-30rc5-gcc46-0.patch -- you can get Intel Core i7 and Intel Core i7 AVX choices in your kernel. Bruce -- Happy Penguin Computers ') 126 Fenco Drive ( \ Tupelo, MS 38801 ^^ supp...@happypenguincomputers.com 662-269-2706 662-205-6424 http://happypenguincomputers.com/ A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing in e-mail? Don't top-post: http://en.wikipedia.org/wiki/Top_post#Top-posting
[gentoo-user] SSH UseDNS without IPv6?
Hi list! I have an issue with SSH. It's a variation of the old Set 'UseDNS no' to avoid delays with faulty DNS records theme. Following setup: 1. I have a server with IPv6 compiled into the SSH daemon but no actual IPv6 network interface. 2. The SSH client has no IPv6, neither compiled nor active. 3. The DNS server doesn't serve or support records. Apparently it drops all such requests. All other records for IP and reverse lookup are correct. Now I'm experiencing the classic, very long delay when connecting to the server via SSH because it does DNS lookups. When I look at wireshark dumps, I see correctly served A and reverse lookups but the server also insists on doing requests which time out. I tried limiting the sshd AddressFamily to inet (aka IPv4) but this didn't change anything. Is there another workaround or do I really have to deactivate DNS lookups? Thanks in advance! Florian Philipp signature.asc Description: OpenPGP digital signature
Re: [gentoo-user] systemd-197-r1 starts gdm-3.6.2
Next episode: I also migrated my gentoo thinkpad to systemd today. Generally very similar to my desktop ... ~amd64 with Gnome 3.6. Things went pretty well, I have to say. I can login to gdm here! ;-) An issue I haven't solved yet: encrypted swap. I always get timeouts as systemd waits for the decrypted mapper-device to come up. Swap doesn't get enabled but when it finally continues to boot I see the valid mapper-device there and can swapon it manually. # cat /etc/crypttab swap /dev/disk/by-id/ata-INTEL_SSDSA2M080G2GC_CVPO015404LR080JGN-part5 /dev/urandom swap,cipher=aes-cbc-essiv:sha256,size=256 # grep swap /etc/fstab /dev/mapper/swapnoneswapdefaults0 0 AFAI understand these 2 lines should be enough to let systemd generate its relevant unit-files etc. Right? Best regards, have a nice weekend, Stefan
Re: [gentoo-user] systemd-197-r1 starts gdm-3.6.2
Am 2013-02-09 19:56, schrieb Stefan G. Weichinger: AFAI understand these 2 lines should be enough to let systemd generate its relevant unit-files etc. Right? Additional thoughts: Is pam_mount obsolete with systemd? It is possible to mount my /home via systemd-unit as well ... the difference seems to be that systemd would (try to) mount it at boot-time while with pam_mount it would be mounted at login. Thoughts? Experiences? Stefan
Re: [gentoo-user] SSH UseDNS without IPv6?
On 09/02/2013 20:22, Florian Philipp wrote: Hi list! I have an issue with SSH. It's a variation of the old Set 'UseDNS no' to avoid delays with faulty DNS records theme. Following setup: 1. I have a server with IPv6 compiled into the SSH daemon but no actual IPv6 network interface. 2. The SSH client has no IPv6, neither compiled nor active. 3. The DNS server doesn't serve or support records. Apparently it drops all such requests. All other records for IP and reverse lookup are correct. Now I'm experiencing the classic, very long delay when connecting to the server via SSH because it does DNS lookups. When I look at wireshark dumps, I see correctly served A and reverse lookups but the server also insists on doing requests which time out. When you say the server also insists on doing requests you mean the SSH server, right? I tried limiting the sshd AddressFamily to inet (aka IPv4) but this didn't change anything. Is there another workaround or do I really have to deactivate DNS lookups? Is the server Gentoo and do you really need IPv6 support on it? Did you consider rebuilding that host with IPv6 disabled in USE? IPv6 coexisting with IPv4 is always going to be a tricky problem, and the recommended defaults you run into all over are usually intended to force people to hurry IPv6 implementation along :-) There's always a way to change defaults, and I found this: http://askubuntu.com/questions/32298/prefer-a-ipv4-dns-lookups-before-ipv6-lookups The magic file you need to edit appears to be /etc/gai.conf -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] Creating accounts in Thunderbird
On 07/02/2013 23:07, Tanstaafl wrote: Which is silly, as username+hostname is not guaranteed to be a singleton in any universe. ? I can't think of any way that username+incoming-hostname can result in anything other than a single, individual users account, so I guess I'm totally missing what you are saying. it A few examples off the top of my head: 1. Two imap servers on the same host running on different ports and no reason why a user can't have accounts on both servers 2. port forwarding on localhost to a variety of impa servers somewhere else (port forwarding gets around corporate firewall rules that Thunderbird can't deal with) 3. Because I can and there's no legitimate reason for a mail client to get in my way 4. Corporate sysadmins like me use tricks like this all the time to a) fix real problems b) comply with frantic business requests c) stay within budget d) get around stupid rules proclaimed by idiot managers with single figure IQs There are more valid reasons why this setup can occur and I have a lack of mentions in RFCs to prove it. There are no valid reasons for a mail client to get in my way like this and I have a lack of RFC mentions that allow it to prove -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] Re: multiple installs
On Sat, 2013-02-09 at 21:20 +1100, Adam Carter wrote: You could install to one of the Athlons, then copy it to the other two Athlons and one of the FX machines. Then, reconfigure the FX install for more CPUs and native -march, recompile world and copy it to the other two FX machines. Even that might not be worth it. Unless you're using something that benefits from the new instructions in the FXes, or requires the last .1% performance, a single install on an Athlon and copy onto the remaining 5 systems will give you more time for something else. From a support point of view the consistency can be a win. Just remember to specify the Althlon64 CPU type in the kernel config, but with 8 cores max. For the hardware, just select a superset of the two systems requirements in your kernel config. If you build as modules any superfluous code wont even be loaded. That might even depend on the compiler version anyway. But, I agree. Building with -march=athlon (I think it's athlon?) on one FX machine and copying to the rest would be fastest. -- Alecks Gates aleck...@gmail.com
Re: [gentoo-user] systemd-197-r1 starts gdm-3.6.2
On Sat, Feb 9, 2013 at 12:56 PM, Stefan G. Weichinger li...@xunil.at wrote: Next episode: I also migrated my gentoo thinkpad to systemd today. Cool. Generally very similar to my desktop ... ~amd64 with Gnome 3.6. Things went pretty well, I have to say. I can login to gdm here! ;-) Try to list the differences between your laptop and your desktop (world files, USE flags, partition schemes, etc.) That was my approach when you described your problem to me; try to see what it differentiates from mine, but we never got too far with tat. An issue I haven't solved yet: encrypted swap. I always get timeouts as systemd waits for the decrypted mapper-device to come up. Swap doesn't get enabled but when it finally continues to boot I see the valid mapper-device there and can swapon it manually. # cat /etc/crypttab swap /dev/disk/by-id/ata-INTEL_SSDSA2M080G2GC_CVPO015404LR080JGN-part5 /dev/urandom swap,cipher=aes-cbc-essiv:sha256,size=256 # grep swap /etc/fstab /dev/mapper/swapnoneswapdefaults0 0 AFAI understand these 2 lines should be enough to let systemd generate its relevant unit-files etc. Right? I haven't used an encrypted swap (nor partition), but I believe that's all you need. A workaround perhaps is to put the nofail option, which at least will skip the partition when booting. Regards. -- Canek Peláez Valdés Posgrado en Ciencia e Ingeniería de la Computación Universidad Nacional Autónoma de México
Re: [gentoo-user] systemd-197-r1 starts gdm-3.6.2
On Sat, Feb 9, 2013 at 1:44 PM, Stefan G. Weichinger li...@xunil.at wrote: Am 2013-02-09 19:56, schrieb Stefan G. Weichinger: AFAI understand these 2 lines should be enough to let systemd generate its relevant unit-files etc. Right? Additional thoughts: Is pam_mount obsolete with systemd? I don't know if obsolete is the correct definition, but it is not installed in any of my systems. It is possible to mount my /home via systemd-unit as well ... the difference seems to be that systemd would (try to) mount it at boot-time while with pam_mount it would be mounted at login. You can mount almost all partitions with system units; there was a discussion some days ago about getting rid of /etc/fstab for the embedded case and stuff like that. Also, you can set the .mount unit for your $HOME, and make the gdm service depend on it (it would be mounted at gdm startup, not at session startup, though). Thoughts? Experiences? I have never used pam_mount; what's the upside? Just delaying the mounting (and perhaps fsck'ing) of the partition until session login? Regards. -- Canek Peláez Valdés Posgrado en Ciencia e Ingeniería de la Computación Universidad Nacional Autónoma de México
Re: [gentoo-user] {OT} LWP::UserAgent slows website
Sure, so long as Apache doesn't have any additional modules loaded. If it's got something like mod_php loaded (extraordinarily common), mod_perl or mod_python (less common, now) then the init time of mod_php gets added to the init time for every request handler. Interesting, so if you have to use mod_php you'd probably be better off running Worker than Prefork, and you'd want to keep MaxConnectionsPerChild on the higher side, to reduce init work you've mentioned, right? May also help to verify that KeepAlive is on and tweak MaxKeepAliveRequests a little higher.
Re: [gentoo-user] Re: multiple installs
That might even depend on the compiler version anyway. But, I agree.
Building with -march=athlon (I think it's athlon?) on one FX machine and
copying to the rest would be fastest.
Good call, I missed that. Of course it makes much more sense to build on a
fast FX box with -j8, but have CFLAGs for the X2, which according to the
safe cflags wiki (
http://en.gentoo-wiki.com/wiki/Safe_Cflags/AMD#Athlon_64_X2) is:
CHOST=x86_64-pc-linux-gnu
CFLAGS=-march=k8 -O2 -pipe
CXXFLAGS=${CFLAGS}
Re: [gentoo-user] {OT} LWP::UserAgent slows website
On Feb 9, 2013 9:26 PM, Adam Carter adamcart...@gmail.com wrote: Sure, so long as Apache doesn't have any additional modules loaded. If it's got something like mod_php loaded (extraordinarily common), mod_perl or mod_python (less common, now) then the init time of mod_php gets added to the init time for every request handler. Interesting, so if you have to use mod_php you'd probably be better off running Worker than Prefork, and you'd want to keep MaxConnectionsPerChild on the higher side, to reduce init work you've mentioned, right? May also help to verify that KeepAlive is on and tweak MaxKeepAliveRequests a little higher. Can't; mod_php isn't compatible with mpm_worker. You have to use a single-threaded mpm like prefork or itk. Anyway, you're starting to get the idea why you want a caching proxy in front of apache.
Re: [gentoo-user] {OT} LWP::UserAgent slows website
Can't; mod_php isn't compatible with mpm_worker. You have to use a single-threaded mpm like prefork or itk. Anyway, you're starting to get the idea why you want a caching proxy in front of apache. Indeed. Thanks for your comments.
Re: [gentoo-user] {OT} LWP::UserAgent slows website
The responses all come back successfully within a few seconds. Can you give me a really general description of the sort of problem that could behave like this? Your server is just a single computer, running multiple processes. Each request from a user (be it you or someone else) requires a certain amount of resources while it's executing. If there aren't enough resources, some of the requests will have to wait until enough others have finished in order for the resources to be freed up. Here's where I'm confused. The requests are made via a browser and the response is displayed in the browser. There is no additional processing besides the display of the response. The responses are received and displayed within about 3 seconds of when the requests are made. Shouldn't this mean that all processing related to these transactions is completed within 3 seconds? If so, I don't understand why apache2 seems to bog down a bit for about 10 minutes afterward. - Grant To really simplify things, let's say your server has a single CPU core, the queries made against it only require CPU consumption, not disk consumption, and the queries you're making require 3s of CPU time. If you make a query, the server will spend 3s thinking before it spits a result back to you. During this time, it can't think about anything else...if it does, the server will take as much longer to respond to you as it takes thinking about other things. Let's say you make two queries at the same time. Each requires 3s of CPU time, so you'll need a grand total of 6s to get all your results back. That's fine, you're expecting this. Now let's say you make a query, and someone else makes a query. Each query takes 3s of CPU time. Since the server has 6s worth of work to do, all the users will get their responses by the end of that 6s. Depending on how a variety of factors come into play, user A might see his query come back at the end of 3s, and user B might see his query come back at the end of 6s. Or it might be reversed. Or both users might not see their results until the end of that 6s. It's really not very predictable. The more queries you make, the more work you give the server. If the server has to spend a few seconds' worth of resources, that's a few seconds' worth of resources unavailable to other users. A few seconds for a query against a web server is actually a huge amount of time...a well-tuned application on a well-tuned webserver backed by a well-tuned database should probably respond to the query in under 50ms! This is because there are often many, many users making queries, and each user tends to make many queries at the same time. There are several things you can do to improve the state of things. The first and foremost is to add caching in front of the server, using an accelerator proxy. (i.e. squid running in accelerator mode.) In this way, you have a program which receives the user's request, checks to see if it's a request that it already has a response for, checks whether that response is still valid, and then checks to see whether or not it's permitted to respond on the server's behalf...almost entirely without bothering the main web server. This process is far, far, far faster than having the request hit the serving application's main code. The second thing is to check the web server configuration itself. Does it have enough spare request handlers available? Does it have too many? If there's enough CPU and RAM left over to launch a few more request handlers when the server is under heavy load, it might be a good idea to allow it to do just that. The third thing to do is to tune the database itself. MySQL in particular ships with horrible default settings that typically limit its performance to far below the hardware you'd normally find it on. Tuning the database requires knowledge of how the database engine works. There's an entire profession dedicated to doing that right... The fourth thing to do is add caching to the application, using things like memcachedb. This may require modifying the application...though if the application has support already, then, well, great. If that's still not enough, there are more things you can do, but you should probably start considering throwing more hardware at the problem...
[gentoo-user] What to do with /var/run?
I received the following ELOG message after an emerge: * One or more symlinks to directories have been preserved in order to * ensure that files installed via these symlinks remain accessible. This * indicates that the mentioned symlink(s) may be obsolete remnants of an * old install, and it may be appropriate to replace a given symlink with * the directory that it points to. * * /var/run Should I change anything? - Grant
[gentoo-user] Shorewall: iptables: No chain/target/match by that name.
I'm getting the following when restarting shorewall: # /etc/init.d/shorewall restart * Stopping firewall ... * Starting firewall ... iptables: No chain/target/match by that name. How can I find out which chain/target/match I need to compile into the kernel? shorewall-init.log does not indicate any problems and I have LOG_VERBOSITY=2 in shorewall.conf which is the maximum. - Grant
Re: [gentoo-user] KDE 4.10 and plasma crashing
Nilesh Govindrajan wrote: Dale, which version of GCC are you using? I'm on 4.7.2, updating KDE now, will report back after a few hours when it's done. -- Nilesh Govindarajan http://nileshgr.com At first, gcc-4.5.4. The second time, gcc-4.6.3. According to the roach report, they have a fix for this. Just waiting for it to travel from the top to the bottom. Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] What to do with /var/run?
Grant emailgr...@gmail.com wrote: I received the following ELOG message after an emerge: * One or more symlinks to directories have been preserved in order to * ensure that files installed via these symlinks remain accessible. This * indicates that the mentioned symlink(s) may be obsolete remnants of an * old install, and it may be appropriate to replace a given symlink with * the directory that it points to. * * /var/run Should I change anything? - Grant If you do shorewall safe-restart (not using the init-script) you see logging on the screen. -- Joost Roeleveld -- Sent from my Android phone with K-9 Mail. Please excuse my brevity.
Re: [gentoo-user] Shorewall: iptables: No chain/target/match by that name.
On 02/10/13 06:19, Grant wrote: I'm getting the following when restarting shorewall: # /etc/init.d/shorewall restart * Stopping firewall ... * Starting firewall ... iptables: No chain/target/match by that name. How can I find out which chain/target/match I need to compile into the kernel? shorewall-init.log does not indicate any problems and I have LOG_VERBOSITY=2 in shorewall.conf which is the maximum. I hade the same problem. Using shorewall trace restart I could figure out which chain/target/match that was missing. Regards. -- Dan Johansson, http://www.dmj.nu *** This message is printed on 100% recycled electrons! *** 0x2FB894AD.asc Description: application/pgp-keys signature.asc Description: OpenPGP digital signature
Re: [gentoo-user] What to do with /var/run?
J. Roeleveld jo...@antarean.org wrote: Grant emailgr...@gmail.com wrote: I received the following ELOG message after an emerge: * One or more symlinks to directories have been preserved in order to * ensure that files installed via these symlinks remain accessible. This * indicates that the mentioned symlink(s) may be obsolete remnants of an * old install, and it may be appropriate to replace a given symlink with * the directory that it points to. * * /var/run Should I change anything? - Grant If you do shorewall safe-restart (not using the init-script) you see logging on the screen. -- Joost Roeleveld Oops... Replied on wrong thread. This was meant for the shorewall one. I blame the mail client on the phone... -- Sent from my Android phone with K-9 Mail. Please excuse my brevity.
