The NetScaler becomes unresponsive because of exhausted swap space.

Reported symptoms include the inability to log on to the system using SSH or through the Admin UI, as well as system crashes. Citrix identified the artifacts necessary to properly diagnose the issue as shown in the following system logs found in the messages file under /var/log directory:

May 27 06:44:24 <kern.crit> pv11p01sa-netscaler001b kernel: swap zone exhausted, increase kern.maxswzone
May 27 06:46:22 <kern.crit> pv11p01sa-netscaler001b kernel: swap zone exhausted, increase kern.maxswzone
May 27 06:46:39 <kern.crit> pv11p01sa-netscaler001b kernel: swap zone exhausted, increase kern.maxswzone
May 27 06:47:05 <kern.crit> pv11p01sa-netscaler001b kernel: swap zone exhausted, increase kern.maxswzone

If you are experiencing this issue, upgrade to a build that meets or exceeds the build numbers for the respective version you are running (see following list). You can download the upgrade from NetScaler Downloads.

Citrix has developed a patch that was introduced in the following builds:

• NetScaler 9.3 build 67.x

• NetScaler 10.1 build 127.x

• All versions of NetScaler 10.5

Problem Cause

NetScaler versions 9.x until 10.1 were developed utilizing FreeBSD 6.3 as the underlying operating system. In the 6.3 version of FreeBSD, the ‘/bin/sh’ executable is vulnerable to a memory leak when performing certain arithmetic expressions. This memory leak is capable of exhausting the system swap space, thereby causing instability in the NetScaler system.

If you are unsure of how to verify this issue, utilize the Customer Insight Services (CIS) system to sign in with your Citrix account and upload a support bundle from the NetScaler to the CIS website system for auto verification.