Re: [SLUG] Increasing RAM
Daniel Pittman wrote: Well, at the moment you have RAID 1 under everything except for swap, which you have on a RAID 0 array. (Implemented, in this case, through the Linux kernel balancing swap space use when areas have equal priority, as your two separate partitions do.) So, how would I perhaps go about setting priority of one SWAP over the other? In that way, based on the fact, I will now have some 6GB RAM in the server, whatever SWAP _is_ being used, I could ensure is most likely only being used in the one space. Thereby increasing the chances (but still having that single POF) that a crash on one disk _may_not_ completely take down the entire system. TiA Kyle -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
Kyle k...@attitia.com writes: Daniel Pittman wrote: Well, at the moment you have RAID 1 under everything except for swap, which you have on a RAID 0 array. (Implemented, in this case, through the Linux kernel balancing swap space use when areas have equal priority, as your two separate partitions do.) So, how would I perhaps go about setting priority of one SWAP over the other? man swapon, find the 'priority' option, and implement in fstab as it specifies. Regards, Daniel -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
On 18/04/2009, at 10:02 PM, Kyle wrote: Hi Slug, I've decided to increase the RAM on my home CentOS server. As best I can recall, the accepted wisdom is to have SWAP approx.~ 2 x RAM. Or was that approx.~ 50% of RAM? Can someone point me in the direction of an explicit tutorial on how I might go about increasing SWAP without destroying data on my other partitions please? Or if I'm actually upping the RAM, should I just not worry about it? These days there's no hard rules about swap. The old rule was 2 x RAM. I don't think you mentioned how much ram you had or how big your swap was, but if you aren't running out of swap, you don't need any more. -- http://chesterton.id.au/blog/ http://barrang.com.au/ -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
RAM is so cheap now, that if you start using swap heavily people just drop in a bit more ! I tend to roughly match swap and memory. At least when i first install. Dean Michael Chesterton wrote: On 18/04/2009, at 10:02 PM, Kyle wrote: Hi Slug, I've decided to increase the RAM on my home CentOS server. As best I can recall, the accepted wisdom is to have SWAP approx.~ 2 x RAM. Or was that approx.~ 50% of RAM? Can someone point me in the direction of an explicit tutorial on how I might go about increasing SWAP without destroying data on my other partitions please? Or if I'm actually upping the RAM, should I just not worry about it? These days there's no hard rules about swap. The old rule was 2 x RAM. I don't think you mentioned how much ram you had or how big your swap was, but if you aren't running out of swap, you don't need any more. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
I used to keep around large swap partitions (that was also before the blissful days of LVM2) until someone on the linux-il mailing list convinced me that the amount of overhead for the kernel to keep track of large amount of swap will actually cause a slow down and reduction of ram utilization. Also remember that the 2X or so rule was from Unix days when the memory management algorithms where totally different (e.g. Each ram page had a shadow swap page pre-allocated for it). So now I stick to around .5 gig, whatever the ram size is. -Amos On 4/18/09, Dean Hamstead d...@fragfest.com.au wrote: RAM is so cheap now, that if you start using swap heavily people just drop in a bit more ! I tend to roughly match swap and memory. At least when i first install. Dean Michael Chesterton wrote: On 18/04/2009, at 10:02 PM, Kyle wrote: Hi Slug, I've decided to increase the RAM on my home CentOS server. As best I can recall, the accepted wisdom is to have SWAP approx.~ 2 x RAM. Or was that approx.~ 50% of RAM? Can someone point me in the direction of an explicit tutorial on how I might go about increasing SWAP without destroying data on my other partitions please? Or if I'm actually upping the RAM, should I just not worry about it? These days there's no hard rules about swap. The old rule was 2 x RAM. I don't think you mentioned how much ram you had or how big your swap was, but if you aren't running out of swap, you don't need any more. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html -- Sent from my mobile device -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
Kyle k...@attitia.com writes: I've decided to increase the RAM on my home CentOS server. As best I can recall, the accepted wisdom is to have SWAP approx.~ 2 x RAM. Or was that approx.~ 50% of RAM? As others have said, this was true back in the days when 64MB was a lot of memory. Now, by the time you are 8GB deep into swap you have spent an awful lot of time thrashing... Can someone point me in the direction of an explicit tutorial on how I might go about increasing SWAP without destroying data on my other partitions please? With fixed partitions, which you seem to use — although you didn't specify the details — life is pretty hard. Do you actually have some unpartitioned space? If not, then forget about repartitioning. If you do, is it next to the existing swap partitions or is it elsewhere? Anyway, if you just want to add swap then swapping to a file is as efficient, these days, as swapping to a raw device.[1] Just use one of them instead if repartitioning is hard. Or if I'm actually upping the RAM, should I just not worry about it? I wouldn't bother. Anyway, a couple of comments: First, this would be vastly easier if you used LVM, since that makes allocating space on the fly a universe easier. Also: Info I'm guessing would be relevant; [k...@bottlenose ~]$ cat /etc/fstab /dev/md1/ ext3defaults1 1 /dev/md2/boot ext3defaults1 2 Are these RAID 0 or RAID 1? If they are RAID 1 then this ... LABEL=SWAP-sdb2 swapswapdefaults0 0 LABEL=SWAP-sda2 swapswapdefaults0 0 ... means that your system will fail when a disk goes bad; you probably want to match the RAID setup underneath your swap to the setup underneath your data devices. Regards, Daniel Footnotes: [1] Technically, since a raw device is only ever one extent and a file may be several it is a few hundred bytes more efficient, I suppose. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
Amos Shapira amos.shap...@gmail.com writes: I used to keep around large swap partitions (that was also before the blissful days of LVM2) until someone on the linux-il mailing list convinced me that the amount of overhead for the kernel to keep track of large amount of swap will actually cause a slow down and reduction of ram utilization. You were, I fear, convinced of something untrue. There is no substantial overhead to tracking a large amount of swap, and certainly no reduction of ram utilization. Linux memory management is not /that/ bad, honest. ;) What they may have been talking about is that having swapped two or three gigabytes of memory usually implies that you will be thrashing, where swapping a few tens or hundreds of megabytes would, historically, be less certain. Also remember that the 2X or so rule was from Unix days when the memory management algorithms where totally different (e.g. Each ram page had a shadow swap page pre-allocated for it). Actually, this has varied across Unix implementations for a range of reasons; under Linux it was historically that way. So now I stick to around .5 gig, whatever the ram size is. *nod* As do I, save in the one case where I want to be able to use the suspend to disk functions supported by Linux. Regards, Daniel -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
First, this would be vastly easier if you used LVM, since that makes allocating space on the fly a universe easier. Re LVM; Maybe, but the last machine I had I ran LVM. I had a hard enough time remembering which volume belonged to which group belonged to which disk (and that despite naming them along the lines of; 'lv00Grp00Hda1', 'lv01Grp00Hda1'). So this time I figured I'd simplify my life somewhat. I have all the necessary conf and data files, etc regularly backed up so if it does go down, it shouldn't be TOO MUCH hassle to get it all back. It is RAID 1. And if I understand Daniel correctly; ... you probably want to match the RAID setup underneath your swap to the setup underneath your data devices. .. You mean I should have the swap spread across the RAID as well. I'm pretty certain that how it is as the m'board controls the RAID of the 2 disks. I just installed on the one array for the OS. Either way, it's too late now. In short, on the advice of all, I just wont bother with increasing SWAP. I'll just dump in the RAM and see what happens. What's bugging me more now is trying to get bloody Ubuntu to recognise and operate my old Linksys-Broadcom Wireless PCI card. But that's a whole nother story. Kindc Regards Kyle Daniel Pittman wrote: Kyle k...@attitia.com writes: [k...@bottlenose ~]$ cat /etc/fstab /dev/md1/ ext3defaults1 1 /dev/md2/boot ext3defaults1 2 Are these RAID 0 or RAID 1? If they are RAID 1 then this ... LABEL=SWAP-sdb2 swapswapdefaults0 0 LABEL=SWAP-sda2 swapswapdefaults0 0 ... means that your system will fail when a disk goes bad; you probably want to match the RAID setup underneath your swap to the setup underneath your data devices. Regards, Daniel -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
On Sunday 19 April 2009 00:16:35 slug-requ...@slug.org.au wrote: I've decided to increase the RAM on my home CentOS server. As best I can recall, the accepted wisdom is to have SWAP approx.~ 2 x RAM. Or was that approx.~ 50% of RAM? Can someone point me in the direction of an explicit tutorial on how I might go about increasing SWAP without destroying data on my other partitions please? Or if I'm actually upping the RAM, should I just not worry about it? Info I'm guessing would be relevant; Of course this is cockamany, urban myth, etc and typically you increase RAM and need even less swap than before for eg (my desktop) Mem: 4036296k total, 4014656k used,21640k free,96400k buffers Swap: 1004020k total, 32k used EXCEPT for 1 tragic circumstance: Never *suspend* unless you have as much SWAP as RAM. Suspend writes all RAM starting at the beginning of swap and over everything along the way. (and I activated suspend by dropping an earring on the keyboard :-) So if your use scenario has never used all your swap, add RAM and forget about adding swap otherwise do the math James -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
Kyle k...@attitia.com writes: First, this would be vastly easier if you used LVM, since that makes allocating space on the fly a universe easier. Re LVM; Maybe, but the last machine I had I ran LVM. I had a hard enough time remembering which volume belonged to which group belonged to which disk (and that despite naming them along the lines of; 'lv00Grp00Hda1', lv01Grp00Hda1'). My immediate response to that is why would you bother? The only case I can think of where it would matter are wanting to have specific partitions on specific disk sets, for performance, which you can rediscover fairly easily if you actually need to. I also note that you don't seem to have enough disks in this machine to make that a concern. Presumably, though, I have missed something, since you obviously had reasons for wanting to know that information. Could you explain them, because I would like to understand what I, um, don't understand. :) So this time I figured I'd simplify my life somewhat. I have all the necessary conf and data files, etc regularly backed up so if it does go down, it shouldn't be TOO MUCH hassle to get it all back. LVM is only a convenience saver, not a backup of any sort, so this would be good practice anyway. It is RAID 1. And if I understand Daniel correctly; ... you probably want to match the RAID setup underneath your swap to the setup underneath your data devices. .. You mean I should have the swap spread across the RAID as well. Well, at the moment you have RAID 1 under everything except for swap, which you have on a RAID 0 array. (Implemented, in this case, through the Linux kernel balancing swap space use when areas have equal priority, as your two separate partitions do.) That also means that when a disk fails it will potentially take out a block of swap with no recourse, and that you will have to manually swap off the physical device to replace it. I'm pretty certain that how it is as the m'board controls the RAID of the 2 disks. I just installed on the one array for the OS. Either way, it's too late now. It isn't, really: you could just turn those two swap partitions into a software RAID array like the rest of your data, but RAID 1 rather than RAID 0. :) In short, on the advice of all, I just wont bother with increasing SWAP. I'll just dump in the RAM and see what happens. That shouldn't be a big drama. What's bugging me more now is trying to get bloody Ubuntu to recognise and operate my old Linksys-Broadcom Wireless PCI card. But that's a whole nother story. You might want to post the details somewhere, see if someone can help. :) Regards, Daniel -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
jam j...@tigger.ws writes: On Sunday 19 April 2009 00:16:35 slug-requ...@slug.org.au wrote: I've decided to increase the RAM on my home CentOS server. As best I can recall, the accepted wisdom is to have SWAP approx.~ 2 x RAM. Or was that approx.~ 50% of RAM? Can someone point me in the direction of an explicit tutorial on how I might go about increasing SWAP without destroying data on my other partitions please? Or if I'm actually upping the RAM, should I just not worry about it? Info I'm guessing would be relevant; Of course this is cockamany, urban myth, etc and typically you increase RAM and need even less swap than before Actually, back in the day this was a good and solid guide, both for performance and safety reasons. Today, less so, but I don't think it is quite as laughable or untrue as you suggest. [...] EXCEPT for 1 tragic circumstance: Never *suspend* unless you have as much SWAP as RAM. You mean suspend to disk, not suspend to RAM, right? Swap is irrelevant to the later, and the amount you need varies with which of the three implementations of the former you choose. However, all of them require as much swap as you have *active memory*, not as much as RAM — although, obviously, if you have no discardable pages[1] then you need the two to be equal. Suspend writes all RAM starting at the beginning of swap and over everything along the way. No, it doesn't. It uses the swap storage space just like the normal kernel, except for adding some private accounting information and a different header to make it possible to detect that it was used to suspend to disk. If it behaved as you describe[2] then it would corrupt memory on the way through as it overwrote swapped data (and, then, no one would ever report a successful suspend to disk. :) Regards, Daniel Footnotes: [1] You do, in the form of all executable code, disk backed memory mappings, disk cache and so forth. In fact, it is quite rare that discardable memory doesn't represent a substantial portion of your used memory, even if swap is in use. [2] Excluding the obvious circumstance of bugs in the suspend software that cause it to create serious corruption, which you may have encountered, especially with the more experimental versions. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
Daniel Pittman wrote: Kyle k...@attitia.com writes: Maybe, but the last machine I had I ran LVM. I had a hard enough time remembering which volume belonged to which group belonged to which disk (and that despite naming them along the lines of; 'lv00Grp00Hda1', lv01Grp00Hda1'). My immediate response to that is why would you bother? Being anal. Plus I was new to LVM at the time. The only case I can think of where it would matter are wanting to have specific partitions on specific disk sets, for performance, exactly. That and local backups. Well, at the moment you have RAID 1 under everything except for swap, which you have on a RAID 0 array. (Implemented, in this case, through the Linux kernel balancing swap space use when areas have equal priority, as your two separate partitions do.) Hmmm. If I think about it... My logic at the time would probably have been; If one disk in the array fails, then all data is mirrored. Yes, the machine might crash and if it had anything in SWAP at that time, I would lose that information. Acceptable risk. However, with a single mirrored disk, and still a complete SWAP partition, I expected I would be able to restart and function on the one disk temporarily until such point in time as I was to rebuild the mirror. Are you saying that wouldn't work? It isn't, really: you could just turn those two swap partitions into a software RAID array like the rest of your data, but RAID 1 rather than RAID 0. :) Ok. So how do I do that? But are you sure my logic above isn't sound? TiA. K -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
On Sunday 19 April 2009 10:00:03 slug-requ...@slug.org.au wrote: On Sunday 19 April 2009 00:16:35 slug-requ...@slug.org.au wrote: I've decided to increase the RAM on my home CentOS server. As best I can recall, the accepted wisdom is to have SWAP approx.~ 2 x RAM. Or was that approx.~ 50% of RAM? Can someone point me in the direction of an explicit tutorial on how I might go about increasing SWAP without destroying data on my other partitions please? Or if I'm actually upping the RAM, should I just not worry about it? Info I'm guessing would be relevant; Of course this is cockamany, urban myth, etc and typically you increase RAM and need even less swap than before Actually, back in the day this was a good and solid guide, both for performance and safety reasons. Today, less so, but I don't think it is quite as laughable or untrue as you suggest. From the days of my first system (PDP11, 100K RAM, 15MB disk) till today I cannot see why this opinion is held. I first encounted it as a RedHat recommendation. Pray wax lyrical ... EXCEPT for 1 tragic circumstance: Never *suspend* unless you have as much SWAP as RAM. You mean suspend to disk, not suspend to RAM, right? Swap is irrelevant to the later, and the amount you need varies with which of the three implementations of the former you choose. However, all of them require as much swap as you have *active memory*, not as much as RAM — although, obviously, if you have no discardable pages[1] then you need the two to be equal. If your active RAM is not equal to physical RAM then the systems is not doing ir right (your definition of active ram ?) Suspend writes all RAM starting at the beginning of swap and over everything along the way. No, it doesn't. It uses the swap storage space just like the normal kernel, except for adding some private accounting information and a different header to make it possible to detect that it was used to suspend to disk. OK it starts SOMEWHERE in swap then writes over everything. In any event I lost my home partition (root swap home) If it behaved as you describe[2] then it would corrupt memory on the way through as it overwrote swapped data (and, then, no one would ever report a successful suspend to disk. :) I never use any suspend and clearly don't appreciate the fine detail, but how would this ever work 1G RAM 2G swap VasttExtravagentApp using 1.9G of swap, then suspend-to-disk? James -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
jam j...@tigger.ws writes: On Sunday 19 April 2009 10:00:03 slug-requ...@slug.org.au wrote: On Sunday 19 April 2009 00:16:35 slug-requ...@slug.org.au wrote: I've decided to increase the RAM on my home CentOS server. As best I can recall, the accepted wisdom is to have SWAP approx.~ 2 x RAM. Or was that approx.~ 50% of RAM? Can someone point me in the direction of an explicit tutorial on how I might go about increasing SWAP without destroying data on my other partitions please? Or if I'm actually upping the RAM, should I just not worry about it? Info I'm guessing would be relevant; Of course this is cockamany, urban myth, etc and typically you increase RAM and need even less swap than before Actually, back in the day this was a good and solid guide, both for performance and safety reasons. Today, less so, but I don't think it is quite as laughable or untrue as you suggest. From the days of my first system (PDP11, 100K RAM, 15MB disk) till today I cannot see why this opinion is held. I first encounted it as a RedHat recommendation. Pray wax lyrical ... Sure. First of all, some of this wisdom comes from other Unix kernels, under which virtual memory management was handled quite differently, and in which it was necessary for the system to preallocate backing store in swap for all pages in use. Second, it also comes from back in the 2.0 through 2.4 era, before the big MM rewrite, under which allocating sufficient swap was significantly helpful to various operations — it meant that the kernel could manage memory more efficiently. Finally, back when a machine with 32MB of RAM was big it was quite conceivable that you would run multiple activities with a combined working set much larger than available memory. Allowing those to swap deeply, when they were not concurrent, meant that you could handle a good deal more context with reasonable performance. Sure, an ideal system wouldn't swap, but a common one would — especially at the hobbiest end of the market. Oh, and it is still possible today to set a strict overcommit mode under Linux, ensuring that every page of allocated anonymous memory has allocated a page of backing store in swap — this works well for ensuring that applications never receive a late out of memory error due to other applications stealing the overcommitted pages from them, at the cost of working poorly with applications that allocate vast chunks of unused virtual memory. EXCEPT for 1 tragic circumstance: Never *suspend* unless you have as much SWAP as RAM. You mean suspend to disk, not suspend to RAM, right? Swap is irrelevant to the later, and the amount you need varies with which of the three implementations of the former you choose. However, all of them require as much swap as you have *active memory*, not as much as RAM — although, obviously, if you have no discardable pages[1] then you need the two to be equal. If your active RAM is not equal to physical RAM then the systems is not doing ir right (your definition of active ram ?) Pages that are not discardable, mostly consisting of anonymous memory. Unmodified page cache, for example, is valuable but not active in this sense. Perhaps not the best choice of term on my part, though. :) Suspend writes all RAM starting at the beginning of swap and over everything along the way. No, it doesn't. It uses the swap storage space just like the normal kernel, except for adding some private accounting information and a different header to make it possible to detect that it was used to suspend to disk. OK it starts SOMEWHERE in swap then writes over everything. In any event I lost my home partition (root swap home) You found an amazingly serious bug; which variant of suspend was this with (or which distribution and release, from which I can derive that?) Anyway, the suspend code is *supposed* to use the standard interface to the swap space, or just the swap space directly, and will not run outside of that space in normal use. If it behaved as you describe[2] then it would corrupt memory on the way through as it overwrote swapped data (and, then, no one would ever report a successful suspend to disk. :) I never use any suspend and clearly don't appreciate the fine detail, but how would this ever work 1G RAM 2G swap VasttExtravagentApp using 1.9G of swap, then suspend-to-disk? ... think, think, think, fail the suspend process with an error because sufficient swap space cannot be found. In other words: this is an error condition, under which normal error reporting should happen and the suspend will not run. The general algorithm of all the suspend implementations is to discard pages until suspend can happen — any discardable pages without concern for swsusp and uswsusp, and as few as possible for TuxOnIce. Then, write what remains to swap, then shut down. If we can't fit everything in swap handle the error gracefully. Regards, Daniel -- SLUG -
Re: [SLUG] Increasing RAM
Kyle k...@attitia.com writes: Daniel Pittman wrote: Kyle k...@attitia.com writes: Maybe, but the last machine I had I ran LVM. I had a hard enough time remembering which volume belonged to which group belonged to which disk (and that despite naming them along the lines of; 'lv00Grp00Hda1', lv01Grp00Hda1'). My immediate response to that is why would you bother? Being anal. Plus I was new to LVM at the time. *nod* OK, that helps explain why. :) The only case I can think of where it would matter are wanting to have specific partitions on specific disk sets, for performance, exactly. That and local backups. Mmmm. Local backups? As in, add another disk and use it as the backup target? That is reasonable. Well, at the moment you have RAID 1 under everything except for swap, which you have on a RAID 0 array. (Implemented, in this case, through the Linux kernel balancing swap space use when areas have equal priority, as your two separate partitions do.) Hmmm. If I think about it... My logic at the time would probably have been; If one disk in the array fails, then all data is mirrored. Yes, the machine might crash and if it had anything in SWAP at that time, I would lose that information. Acceptable risk. However, with a single mirrored disk, and still a complete SWAP partition, I expected I would be able to restart and function on the one disk temporarily until such point in time as I was to rebuild the mirror. Are you saying that wouldn't work? No, that will work as you state: a crash will potentially take your system down, recovery will be fine. If you have accepted that trade-off then all is well. A surprising number of people /don't/ consider that, and can't afford to have the system go down, but still use RAID-0 swap. :) Regards, Daniel -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Increasing RAM
jam == jam j...@tigger.ws writes: jam On Sunday 19 April 2009 10:00:03 slug-requ...@slug.org.au wrote: On Sunday 19 April 2009 00:16:35 slug-requ...@slug.org.au wrote: I've decided to increase the RAM on my home CentOS server. As best I can recall, the accepted wisdom is to have SWAP approx.~ 2 x RAM. Or was that approx.~ 50% of RAM? Can someone point me in the direction of an explicit tutorial on how I might go about increasing SWAP without destroying data on my other partitions please? Or if I'm actually upping the RAM, should I just not worry about it? Info I'm guessing would be relevant; Of course this is cockamany, urban myth, etc and typically you increase RAM and need even less swap than before Actually, back in the day this was a good and solid guide, both for performance and safety reasons. Today, less so, but I don't think it is quite as laughable or untrue as you suggest. jam From the days of my first system (PDP11, 100K RAM, 15MB disk) jam till today I cannot see why this opinion is held. I first jam encounted it as a RedHat recommendation. Pray wax lyrical ... Standard (i.e., Edition 7 and later) unices used to write all their memory into the swap partition on crash, along with some headers. Then when you booted the system, the early code would recognise that the swap partition was full of a memory image, and offer to dump it to tape (or, for Kodak Unix, to 8 floppies!). Then after you'd booted and fscked, you could load the image into a file, and use the crash(8) utility on it to find out why the kernel crashed. For all this to work, you needed a bit more swap than you had RAM --- and two times would always be enough (and it was a good enough rule of thumb that made its way into the SCO system administrator's course --- in the days when machine prices started coming down, and instead of a central system, companies were putting in small servers here there and everywhere; quite often the admin assistant or receptionist would be sent on a course to learn how to set up or administer the thing. I remember giving courses in 1986--8 and they were full of women who'd been told off to learn how to use and administer `the computer'. I taught them vi and troff (word processing) how to install, partition, and format discs, add users, use chmod, use printers, set up serial ports for terminals, etc (all the admin stuff) in three days!!! At least they could generally type. Niceties such as working out exactly how much swap you needed were way too complex, so we told them, twice your RAM) Peter C -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
