Memory Error

[Ricky_2]

Hi, I get the following errors when I logged into my server. I noticed that the memory utilisation is high (100%) but the swap is still alright (15%) from 'swapinfo -a'. What does the error mean and is there something that I can do about it? Thanks.

/usr/lib/dld.sl:Call to mmap () failed - DATA /usr/lib/libxti.2
/usr/lib/dld.sl:Not enough space
/etc/profile[105]:29086 Abort

[Ravi_8]

Hi, ricky

lemme know the RAM size, If the sawp is already diuble of RAM make it thrice (swap).
If you still have the problem, increase maxdsiz and maxdsiz_64bit kernel parameters values.

[Ricky_2]

Hi Ravi,
I have 2GB ram. My maxdsiz and maxdsiz_64bit is 67108864 and 1073741824 respectively. How do I arrive at a new value and what are the implications? Thanks.

[Clemens van Everdingen]

Hi,

After which change did this start ?

Patch install/ sw upgrade/ OS upgrade ?

C.

[Ricky_2]

Hi, it actually came about with a temporary increase in users on the Oracle DB running on my server. I assume that a hardware upgrade is probably the best move but I would like to tune it first to optimise the performance. Also, I hope to find out the role of the /usr/lib/dld.sl. Thanks.

[Paula J Frazer-Campbell]

Ricky

Have you make any kernel changes?

Paula

[Ravi_8 · Accepted solution]

Hi,

At the very least you need to increase maxdsiz (or maxdsiz_64 if this is a 64-bit application.) What has happened, is that recently a memory-mapped file (I suspect a very heavily used table or the dictionary itself in the database) has finally grown large enough to exceed the process memory limit (maxdsiz). Increase it and generate a new kernel and that should get you going. You probably need to look at adding more memory to your system as well but maxdsiz is your immediate problem.


'maxdsiz' (for 32-bit machines) has a minimum value of 0x400000 (4MB), a maximum of 0x7B03A000 (~ 2GB) and a default of 0x4000000 (64MB).

64MB is the default value so 256-512MB should do the trick

[Ricky_2]

Hi Paula, I have not done anything to the kernel. I would like to find out more about the cause and options before I make the changes. Thanks.

[Paula J Frazer-Campbell]

Hi

The problem is most likely related to a kernel parameter being to low.

Run :-

sar -v 1 100

whilst this problem is occurring and note any table overflows. also have a look at /var/adm/syslog/syslog.log and note/ post any errors.


Paula