I finally found the issue.
It wasn't a kernel limit, as this operational account always ran the
"unlimit" csh command. This boosted the data size up to 1 gig. The issue
here was in the actual link step of building our application. It was using
some "-taso" flag for the compiler. In any event, it still will be a problem
for me as our application doesn't require this flag, but one of our required
3PP librarys we link must have this. So I am screwed here too. Need to
figure out why that is the case.
Thanks all.
-grant
-----Original Message-----
From: Schoep, Grant _at_ STORM
Sent: Wednesday, June 11, 2003 6:23 PM
To: 'tru64-unix-managers_at_ornl.gov.'
Subject: Process can't allocate more memory, even though there is more
Having an issue with a Tru64 4.0f machine(ok all of them).
We have a process that is growing fairly steadily. It really uses a lot of
memory. Under Solaris it levels off around 700 megs. But on 4.0f, the app
errors out around 300 meg as a malloc call returns 0(out of memory).
I'm guessing this is one of those many really annoying kernel limit settings
that is limiting our process's memory size. I thought I figured it out, but
am having a bit of trouble.
Looking at the Sys Config and Tuning guide, I guessed it was
per-proc-data-size. So I kicked it up 5 times from its default. I just
rebooted after this(this doesn't need a kernel rebuild does it?)
sysconfig -q proc now reports the new number.
However, app still errors out at 300 megs.
So, I ran trusty old sys_check -perf and made all its recommended
changes(and backed out the previous one as it didn't change anything).
No luck... still errors out.
Some info about the machine. It only has 1 gig. But on our normal machines
that all have 2-4 gigs of RAM they do the same thing.
Its a PW 1000XP
It has 3 gigs of swap space.
It is running 4.0f( upgrading to any patch level from this is not currently
an option...)
I reproduced this issue quickly by making some code that basically just runs
in a for loop and mallocs tons of memory until malloc returns 0. It does
this at about 300 megs.
So more machine info below. If anyone can point me something to try that
would be great.
Thanks
-grant
As it may help, I also am pasting in the following info. Note that all our
processes are running after a "unlimit(csh)" command. Output of the limit
command before and after.
grant maroon>27% limit
cputime unlimited
filesize unlimited
datasize 131072 kbytes
stacksize 2048 kbytes
coredumpsize unlimited
memoryuse 1020384 kbytes
descriptors 4096 files
addressspace 1048576 kbytes
grant_at_maroon>28% unlimit
grant_at_maroon>29% limit
cputime unlimited
filesize unlimited
datasize 1048576 kbytes
stacksize 32768 kbytes
coredumpsize unlimited
memoryuse 1019856 kbytes
descriptors 4096 files
addressspace 1048576 kbytes
grant_at_maroon>30% ulimit
I also did a "sysconfig -q" on vm and proc as I figured those are most
applicable. The output of these was AFTER I setup the recommended settings
from syscheck -perf.
rant_at_maroon>27% limit
cputime unlimited
filesize unlimited
datasize 131072 kbytes
stacksize 2048 kbytes
coredumpsize unlimited
memoryuse 1020384 kbytes
descriptors 4096 files
addressspace 1048576 kbytes
grant_at_maroon>28% unlimit
grant_at_maroon>29% limit
cputime unlimited
filesize unlimited
datasize 1048576 kbytes
stacksize 32768 kbytes
coredumpsize unlimited
memoryuse 1019856 kbytes
descriptors 4096 files
addressspace 1048576 kbytes
grant_at_maroon>30% ulimit
root_at_maroon>1% sysconfig -q proc
proc:
max-proc-per-user = 384
max-threads-per-user = 384
per-proc-stack-size = 2097152
max-per-proc-stack-size = 33554432
per-proc-data-size = 134217728
max-per-proc-data-size = 1073741824
max-per-proc-address-space = 1073741824
per-proc-address-space = 1073741824
executable_stack = 0
autonice = 0
autonice-time = 600
autonice-penalty = 4
open-max-soft = 4096
open-max-hard = 4096
ncallout_alloc_size = 8192
round-robin-switch-rate = 0
round_robin_switch_rate = 0
sched-min-idle = 0
sched_min_idle = 0
give-boost = 1
give_boost = 1
maxusers = 512
task-max = 4117
thread-max = 8232
num-wait-queues = 512
num-timeout-hash-queues = 512
enable_extended_uids = 0
enhanced-core-name = 1
enhanced-core-max-versions = 3
vm:
ubc-minpercent = 10
ubc-maxpercent = 100
ubc-borrowpercent = 20
ubc-maxdirtywrites = 5
vm-max-wrpgio-kluster = 32768
vm-max-rdpgio-kluster = 16384
vm-cowfaults = 4
vm-mapentries = 16384
vm-maxvas = 1073741824
vm-maxwire = 16777216
vm-heappercent = 7
vm-vpagemax = 131072
vm-segmentation = 1
vm-ubcpagesteal = 24
vm-ubcdirtypercent = 10
vm-ubcseqstartpercent = 40
vm-ubcseqpercent = 10
vm-csubmapsize = 1048576
vm-ubcbuffers = 256
vm-syncswapbuffers = 128
vm-asyncswapbuffers = 4
vm-clustermap = 1048576
vm-clustersize = 65536
vm-zone_size = 0
vm-kentry_zone_size = 16777216
vm-syswiredpercent = 80
vm-inswappedmin = 1
vm-page-free-target = 256
vm-page-free-swap = 138
vm-page-free-hardswap = 4096
vm-page-free-min = 20
vm-page-free-reserved = 10
vm-page-free-optimal = 138
vm-page-prewrite-target = 512
dump-user-pte-pages = 0
kernel-stack-guard-pages = 1
vm-min-kernel-address = 18446744071562067968
malloc-percpu-cache = 1
vm-aggressive-swap = 0
vm_page_color_private = 0
vm-map-index-count = 64
vm-map-index-rebalance = 128
vm-map-index-enabled = 1
vm-map-index-hiwat = 4
vm-map-index-lowat = 2
new-wire-method = 1
vm-segment-cache-max = 50
vm-page-lock-count = 0
gh-chunks = 0
gh-min-seg-size = 8388608
gh-fail-if-no-mem = 1
private-text = 0
Received on Thu Jun 12 2003 - 16:18:13 NZST