similar to: Identifying anonymous errors, disabling anonymous tracing

Hi, I am profiling some workloads for the voluntary and involuntary context switches. I am interested in finding out the reasons causing these two types of context switches. As far as I understand, involuntary context switch happens on expiration of time slice or when a higher priority process comes in. While the voluntary switch generally happens when a process is waiting for I/O etc. So to

How can I tell if this is normal behaviour? Oracle imports are horribly slow, an order of magnitude slower than on the same hardware with a slower disk array and Solaris 9. What I can look for to see where the problem lies? The server is 99% idle right now, with one database running. Each sample is about 5 seconds. I''ve tried setting kernel parameters despite the docs saying that

Hi, I am fairly new to DTrace, and wanted to ask something regarding synchronization objects. I have a multithreaded C++ program in which I use semaphores (POSIX version) for synchronization. I was interested in using DTrace to figure out the total time for which each thread in my application sleeps, blocked on a semaphore, i.e., as a result of sem_wait() call. In order to measure this, I

I''m looking for an example of how one could write a dtrace probe that could follow something like a NULL terminated linked list. For example: struct list { struct list *next; void *data; }; struct list *list_root; Assuming I had "list_root" available, but I wanted to know how long the linked list is, how can I do this in a dtrace probe? If I had looping

Hi, I have a faulty dtrace script (most probably too big bufsize). My machine hangs at boot (amd64 5.10 Version Generic_118855-10 64-bit). NOTICE: enabling probe 11 (dtrace:::ERROR) WARNING: /etc/svc/volatile: File system full, swap space limit exceeded WARNING: Sorry, no swap space to grow stack for pid 5 (autopush) WARNING: /etc/svc/volatile: File system full, swap space limit exceeded How to

Hello experts, I heard that there is something called anonymous dtrace that would still be running when I do a reboot. Basically, I have the following problem: The /boot/solaris/bootenv.rc file in my alternate boot environment is getting modified when I reboot the machine after doing luactivate <ABE>. It happens only on init 6, doesn''t happen when I do a simple reboot. The set

Hello, I''ve done this several times in the past but a co-worker is having problems getting anonymous tracing to work: dtrace -A -m su and we see updated /kernel/drv/dtrace.conf and fbt and dtrace forceloads added to /etc/system. Upon reboot the following errors occur (note that we are running snv_66): {0} ok boot disk Boot device: /pci at 0/pci at 0/pci at 2/scsi at 0/disk at 0 File

Hi there, I am trying to diagnose a performance problem on a Sun T5120 running Solaris 10. I see general poor performance using Java and, in particular, loading classes. I have developed a simple java test that loads all of the classes in a Crystal Reports JAR file (around 7000 classes) which takes much longer on the T5120 (40 secs) than on a Core 2 Duo windows PC (5 secs). I recognise that this

I''m using DTrace''s anonymous tracing feature to trace through device drivers at boot time. With Solaris 10 Update 6 B5, I was unable to trace fbt based calls from the sd attach point. The error messages from the console are: NOTICE: enabling probe 0 (fbt::sdattach:entry) NOTICE: enabling probe 1 (fbt:::) NOTICE: enabling probe 2 (fbt::sdattach:return) NOTICE: enabling probe

Hi, Has something related to speculative tracing changed between s10 FCS and the more recent nevada builds ? I was trying the specopen.d script from the Dtrace guide on a nevada machine and it failed with : dtrace: failed to enable ''./spec.d'': DIF program content is invalid To try and narrow things down a bit I wrote the following short script. This works fine on s10, but

[1] Predicates in one-liners I would like to list the probe modules in my executable and then dynamically create a dscript to trace execution of those modules alone (by excluding the 3rd party and system libraries). I tried the below script without success. The conditional given in the predicate is not taking effect. Why is this so ? $ dtrace -ln ''pid$target:::

I have a sunw2100z (dual AMD) 2GB memory, I refresh installed nvx_35 on scsi disk (on board controller), then I enabled anon tracing fbt::xxx_attach:entry { self->trace = 1; } fbt::: /self->trace/ {} fbt::xxx_attach:return { self->trace = 0; } xxx_attach is HBA driver''s attach entry point. #dtrace -AFs xxx.d # reboot then reboot gives enabling probe 0..... WARNING:

Hi all, Is there some documentation or some example on how to interpret the arg0 .. arg<n> for the aioread, aiowrite, aiowait syscalls? The system call name for all three seems to be "kaio". Michael === Michael Mueller ================== Tel. + 49 8171 63600 Fax. + 49 8171 63615 Web: http://www.michael-mueller-it.de ======================================

Tracers, Good news! The Solaris Dynamic Tracing Guide has entered the participation age at last and has been converted into wiki format. The new and shiny guide can be found at: http://wikis.sun.com/display/DTrace/Documentation The guide has fallen behind with developments in DTrace since it was first released and needs some revising and some bug fixing as well. However, now that it''s

I''ve recently re-installed an X4500 running Nevada b109 and have been experiencing ZFS lock ups regularly (perhaps once every 2-3 days). The machine is a backup server and receives hourly ZFS snapshots from another thumper - as such, the amount of zfs activity tends to be reasonably high. After about 48 - 72 hours, the file system seems to lock up and I''m unable to do anything

Hello, I wanted to enable tracing in user process. but following example from guide is not working. pid$1::$2:entry { self->trace = 1; } pid$1::$2:return /self->trace/ { self->trace = 0; } pid$1:::entry, pid$1:::return /self->trace/ { } I am getting error like invalid probe description ..... undefined macro variable in probe description. pid$target:user lib:user function:entry

Hi everyone, I would like to identify the case by groups that is just bigger that avg plus sd. For example, using species as group and petal.wid as my variable in the iris data. What's the better way to doit? creating a function? So,the question is to identify the single element of each species that is just larger than a cut-off point (i.e. larger than mean + sd) I made this, but I can not

Hi, I know how to trace the number of bytes read/written by a program by summing the argument of that read/write syscall. However, the way the OS services read/write calls is in blocks. I wonder if it''s possible to trace the actual number of blocks read/written caused by the syscalls? -- This message posted from opensolaris.org

Hello everyone, I want make a variable selection  from a dataframe, but when I build this new object (using cbind) the new table (which is a matrix) I lost all the original factor names in the variables. I get 1,2,3.... Someone would be so kind and tell me if there's is a ny way to get variables in a dataframe, but not changing these factor to numbers? Thank you all. José [[alternative

We are using following dtrace script to analyze an application. However we see different number of lwp_yield and yield calls reported depending on whether we trace mutex_lock or mutex_unlock or not. The number reported when we are not tracing mutex_lock or mutex_unlock is higher. What could be going on here and which one is the correct number. This is on S10 U2. Thanks, Rao.