Re: 3.9 freeze
Pedro Martelletto wrote: I cannot declare that the problem is solved... but I had no more freezes since I'm using a custom GENERIC kernel with doubled NKMEMPAGES_MAX and maxusers, both with the i386 and the amd64 machines. But consider that this happened only 7 and 10 days ago... It has been approximately a month now. How have your boxes been doing? They both (i386 and amd64) had not a single freeze any more. Bye. -- ___ __ |- [EMAIL PROTECTED] |ederico Giannici http://www.neomedia.it ___
Re: 3.9 freeze
I cannot declare that the problem is solved... but I had no more freezes since I'm using a custom GENERIC kernel with doubled NKMEMPAGES_MAX and maxusers, both with the i386 and the amd64 machines. But consider that this happened only 7 and 10 days ago... It has been approximately a month now. How have your boxes been doing? -p.
Re: 3.9 freeze
Pedro, since I set the option NKMEMPAGES_MAX=65535 on kernel file, the server doesn't freeze UVM amap128305 10153K 50705K157284K4071891000 0 16,32,64,128,256,512,1024,2048,4096,8192,16384,32768,65536 This server has an uptime 12 days, before the change only alive 3 or 4 days regards,. - Original Message - From: Pedro Martelletto [EMAIL PROTECTED] To: Federico Giannici [EMAIL PROTECTED] Cc: misc@openbsd.org; diego [EMAIL PROTECTED] Sent: Saturday, July 29, 2006 2:00 PM Subject: Re: 3.9 freeze Any news on this? -p.
Re: 3.9 freeze
Any news on this? -p.
Re: 3.9 freeze
Pedro Martelletto wrote: Any news on this? I cannot declare that the problem is solved... but I had no more freezes since I'm using a custom GENERIC kernel with doubled NKMEMPAGES_MAX and maxusers, both with the i386 and the amd64 machines. But consider that this happened only 7 and 10 days ago... Bye. -- ___ __ |- [EMAIL PROTECTED] |ederico Giannici http://www.neomedia.it ___
Re: 3.9 freeze
Federico Giannici wrote: Pedro Martelletto wrote: Any news on this? I cannot declare that the problem is solved... but I had no more freezes since I'm using a custom GENERIC kernel with doubled NKMEMPAGES_MAX and maxusers, both with the i386 and the amd64 machines. But consider that this happened only 7 and 10 days ago... I have to add that both machines have a lot of memory (2GB) and often a lot of processes running. Bye. -- ___ __ |- [EMAIL PROTECTED] |ederico Giannici http://www.neomedia.it ___
Re: 3.9 freeze
Pedro Martelletto wrote: Federico, Your diagnosis is correct, that freeze can be the result of reaching the limit for UVM amap allocations. These get used by the kernel to describe anonymous memory mappings, and mmap malloc() puts the UVM subsystem under a higher load of those, eventually reaching the limit. Until an appropriate solution is found, you can try bumping the number of pages in the kernel's memory map (NKMEMPAGES). I have just installed a new 3.9-stable kernel with doubled NKMEMPAGES_MAX (and maxusers). I'll keep you informed if this solved the problem... Bye. P.S. It seems to me that under amd64 the UVM Amap usage is much higher then under i386. So, even if by default the limit is the double of the i386, it seems not enough. -- ___ __ |- [EMAIL PROTECTED] |ederico Giannici http://www.neomedia.it ___
Re: 3.9 freeze
On Thu, Jul 20, 2006 at 06:07:16PM +0200, Federico Giannici wrote: It seems to me that under amd64 the UVM Amap usage is much higher then under i386. So, even if by default the limit is the double of the i386, it seems not enough. That's probably because the code allocates in multiples of sizeof(long), which is twice as bigger in 64-bit architectures than on 32-bit. I still have to do the math correctly, but the default limit may even not be enough for some i386 systems with gigantic amounts of memory. Ideally, the kernel should have a smarter way of setting the limit for malloc() in kmeminit(), perhaps based on physmem, or on the ability of growing its memory mapping dynamically, as needed. -p.
Re: 3.9 freeze
Federico, Your diagnosis is correct, that freeze can be the result of reaching the limit for UVM amap allocations. These get used by the kernel to describe anonymous memory mappings, and mmap malloc() puts the UVM subsystem under a higher load of those, eventually reaching the limit. Until an appropriate solution is found, you can try bumping the number of pages in the kernel's memory map (NKMEMPAGES). -p.
Re: 3.9 freeze
Pedro Martelletto wrote: Federico, Your diagnosis is correct, that freeze can be the result of reaching the limit for UVM amap allocations. These get used by the kernel to describe anonymous memory mappings, and mmap malloc() puts the UVM subsystem under a higher load of those, eventually reaching the limit. Until an appropriate solution is found, you can try bumping the number of pages in the kernel's memory map (NKMEMPAGES). I'm not sure of what variables to set and where. Is it correct to add the following line to the kernel configuration file? option NKMEMPAGES_MAX 65536 Thanks. -- ___ __ |- [EMAIL PROTECTED] |ederico Giannici http://www.neomedia.it ___
Re: 3.9 freeze
On Tue, Jul 18, 2006 at 07:34:00PM +0200, Federico Giannici wrote: I'm not sure of what variables to set and where. options(4) should tell you that. -p.
Re: 3.9 freeze
Federico, I put option NKMEMPAGES_MAX=65535 on the kernel config. vmstat -m show that UVM amap 68283 2676K 2871K157284K 2166240 0 16,32,64,128,256,512,1024,2048,4096,32768,65536 the limit now is 157284K, before was 39322K. regards,. - Original Message - From: Federico Giannici [EMAIL PROTECTED] To: Pedro Martelletto [EMAIL PROTECTED] Cc: misc@openbsd.org; diego [EMAIL PROTECTED]; mickey [EMAIL PROTECTED] Sent: Tuesday, July 18, 2006 2:34 PM Subject: Re: 3.9 freeze Pedro Martelletto wrote: Federico, Your diagnosis is correct, that freeze can be the result of reaching the limit for UVM amap allocations. These get used by the kernel to describe anonymous memory mappings, and mmap malloc() puts the UVM subsystem under a higher load of those, eventually reaching the limit. Until an appropriate solution is found, you can try bumping the number of pages in the kernel's memory map (NKMEMPAGES). I'm not sure of what variables to set and where. Is it correct to add the following line to the kernel configuration file? option NKMEMPAGES_MAX 65536 Thanks. -- ___ __ |- [EMAIL PROTECTED] |ederico Giannici http://www.neomedia.it ___
Re: 3.9 freeze
scxspl 128 110083960 11008396 4 4 0 4 0 8 0 namei 1024 198672932 0 1986729326 6 0 6 0 8 0 vnodes 156 262100 101 0 101 101 0 8 0 nchpl 72 13100024 02424 0 8 0 ffsino 168 104415410 0 104412794 109 0 109 109 0 8 0 dino1pl 128 104415410 0 104412794 85 08585 0 8 0 pagedeppl 68 1393010 139293 1 0 1 1 0 8 0 inodedeppl84 7282820 728270 609 608 1 222 0 8 0 newblkpl 32 63302310 6330231 1 1 0 1 0 8 0 bmsafemappl 32 1258270 125821 1 0 1 1 0 8 0 allocdirectpl 68 11135130 11134475654 213 0 8 0 indirdeppl2827999027996 1 0 1 1 0 8 0 allocindirpl 52 52167180 5216717 32798 32797 1 136 0 8 0 freefragpl32 2967820 296750 2 1 1 2 0 8 0 freeblkspl 108 5145070 514505 606 605 1 196 0 8 0 freefilepl28 3085630 3085639595 051 0 8 0 diraddpl 32 4861810 4861741110 110 0 8 0 mkdirpl 2822530022530 1 1 0 1 0 8 0 dirrempl 32 5048240 504820 121 120 178 0 8 0 newdirblkpl 16 1720 172 1 1 0 1 0 8 0 dirhash 1024 5135990 512906 21372 21198 174 485 0 128 0 semapl68300 1 0 1 1 0 8 0 semupl 100 1554990 155498 1 0 1 1 0 8 0 pfrulepl 628 910 1216 21416 0 8 0 pfstatepl284 9291980 921933 61192 519 519 0 715 0 pfosfpen 108 7640 38214 31111 0 8 0 pfosfp28 4160 208 2 0 2 2 0 8 0 rtentpl 108 65950 6554 3 1 2 3 0 8 0 rttmrpl 32 94720 9472 1 1 0 1 0 8 0 tcpcbpl 400 3104210 309935 263 2144949 0 8 0 tcpqepl 16 2069030 206903 1 1 0 1 013 0 sackhlpl 20 2166630 216663 1 1 0 1 0 163 0 synpl184 2447290 2447297877 1 4 0 8 1 plimitpl 15210502010480 2 1 1 2 0 8 0 inpcbpl 216 10490980 104860289612828 0 8 0 In use 14564K, total allocated 32736K; utilization 44.5% thanks... diego,. - Original Message - From: diego [EMAIL PROTECTED] To: mickey [EMAIL PROTECTED] Cc: Pedro Martelletto [EMAIL PROTECTED]; misc@openbsd.org Sent: Monday, July 03, 2006 12:19 PM Subject: Re: 3.9 freeze ok, I have the server on datacenter, when freeze I will try it. - Original Message - From: mickey [EMAIL PROTECTED] To: diego [EMAIL PROTECTED] Cc: Pedro Martelletto [EMAIL PROTECTED]; misc@openbsd.org Sent: Monday, July 03, 2006 9:52 AM Subject: Re: 3.9 freeze On Mon, Jul 03, 2006 at 09:45:22AM -0300, diego wrote: no, I can only ping the server or change tty (ctrl alt fn), but I can't type anything. you should sysctl ddb.console=1 for that to work... - Original Message - From: Pedro Martelletto [EMAIL PROTECTED] To: diego [EMAIL PROTECTED] Cc: misc@openbsd.org Sent: Monday, July 03, 2006 9:34 AM Subject: Re: 3.9 freeze Can you break into ddb? -p. -- paranoic mickey (my employers have changed but, the name has remained) [demime 1.01d removed an attachment of type application/octet-stream which had a name of vmstat.bsd.0.core] [demime 1.01d removed an attachment of type application/octet-stream which had a name of dmesg.bsd.0.core] [demime 1.01d removed an attachment of type application/octet-stream which had a name of ps.bsd.0.core]
Re: 3.9 freeze
The next time it freezes, break into ddb and get the output of 'show uvmexp'. -p.
Re: 3.9 freeze
diego wrote: UVM amap201783 39322K 39322K 39322K12379757100 0 16,32,64,128,256,512,1024,2048,4096,8192,16384,32768,65536 I have noticed that the above UVM amap HighUse value is equal to the Limit value. As I have already said, the PC of mine that occasional freezes has high values of VM amap too. So I'm asking again: what happens when the Limit value is reached? Can it make the pc to freeze? What can we do to avoid this? Thanks. -- ___ __ |- [EMAIL PROTECTED] |ederico Giannici http://www.neomedia.it ___
Re: 3.9 freeze
On Thu, Jul 13, 2006 at 05:39:40PM +0200, Federico Giannici wrote: I have noticed that the above UVM amap HighUse value is equal to the Limit value. Indeed it looks suspicious. Not my area, though, so I'd have to look at the code to know the exact consequences. But yes, it's a possibility. Perhaps Mickey or Ted can enlighten us? -p.
Re: 3.9 freeze
Can you break into ddb? -p.
Re: 3.9 freeze
no, I can only ping the server or change tty (ctrl alt fn), but I can't type anything. - Original Message - From: Pedro Martelletto [EMAIL PROTECTED] To: diego [EMAIL PROTECTED] Cc: misc@openbsd.org Sent: Monday, July 03, 2006 9:34 AM Subject: Re: 3.9 freeze Can you break into ddb? -p.
Re: 3.9 freeze
On Mon, Jul 03, 2006 at 09:45:22AM -0300, diego wrote: no, I can only ping the server or change tty (ctrl alt fn), but I can't type anything. you should sysctl ddb.console=1 for that to work... - Original Message - From: Pedro Martelletto [EMAIL PROTECTED] To: diego [EMAIL PROTECTED] Cc: misc@openbsd.org Sent: Monday, July 03, 2006 9:34 AM Subject: Re: 3.9 freeze Can you break into ddb? -p. -- paranoic mickey (my employers have changed but, the name has remained)
Re: 3.9 freeze
no... - Original Message - From: vladas [EMAIL PROTECTED] To: diego [EMAIL PROTECTED] Sent: Monday, July 03, 2006 10:00 AM Subject: Re: 3.9 freeze On 03/07/06, diego [EMAIL PROTECTED] wrote: no, I can only ping the server or change tty (ctrl alt fn), but I can't type anything. how about by ssh? - Original Message - From: Pedro Martelletto [EMAIL PROTECTED] To: diego [EMAIL PROTECTED] Cc: misc@openbsd.org Sent: Monday, July 03, 2006 9:34 AM Subject: Re: 3.9 freeze Can you break into ddb? -p.
Re: 3.9 freeze
ok, I have the server on datacenter, when freeze I will try it. - Original Message - From: mickey [EMAIL PROTECTED] To: diego [EMAIL PROTECTED] Cc: Pedro Martelletto [EMAIL PROTECTED]; misc@openbsd.org Sent: Monday, July 03, 2006 9:52 AM Subject: Re: 3.9 freeze On Mon, Jul 03, 2006 at 09:45:22AM -0300, diego wrote: no, I can only ping the server or change tty (ctrl alt fn), but I can't type anything. you should sysctl ddb.console=1 for that to work... - Original Message - From: Pedro Martelletto [EMAIL PROTECTED] To: diego [EMAIL PROTECTED] Cc: misc@openbsd.org Sent: Monday, July 03, 2006 9:34 AM Subject: Re: 3.9 freeze Can you break into ddb? -p. -- paranoic mickey (my employers have changed but, the name has remained)