Hi Eduardo,

What I can see from the backtrace is that rt_xtt is updating a value field en a Ge graph, and when drawing a line, probably the border line of the field, the error "Fatal IO error 11 (Resource temporarily unavailable) on X server :0.0" is returned from the X server. The X server is the Xorg process so you could have a look at it with top and see if it's eating memory taking a lot of CPU. There is also a log file for Xorg /var/log/Xorg.0.log that might give some more info of what kind of resource is unavailable.

I don't have much experience with problems with the X server but it often has to do with bugs in the driver for the graphic card, so be sure to install the latest version of this.

/Claes

Hi Claes,

Thanks for your comment.

I attached the Xorg.0.log, but I couldn't see anything relevant or related with "Fatal IO error 11 (Resource temporarily unavailable) on X server :0.0" , also in the syslog and xsession-errors.

I updated the nvidia driver and also using Nouveau driver without any change.

It seems that the problem could be related with the Ubuntu12.04LTS.

Are there an inconvinient related to Proview5.0, if I only upgrade Ubuntu from 12.04LTS to 14.04LTS, on both OTStations, leaving the Proview5.0 as it is?

Eduardo

Hi Claes

I want to comment you that, in order to find a solution to the rt_xtt exit due to Fatal I/O error 0, this weekend I installed from scratch the second OTStation with Ubuntu 14.04LTS, pwrtt_5.0.0-1_i386.deb and then the generate and distribute and I found that the rt_qmon is consuming almost 100% of the cpu resources.

Can you also help with this?

Eduardo

Hi Eduardo,

I know that sometimes when an old node suddenly appears in a new version or platform, the nethander gets confused and you have to stop both the op station and the process stations before you start them again.

/Claes

Hi Claes,

No way.

I stopped all the system, 2 Ots, 1 Process Station and the History Server, then I restarted the system, and each time the "new OT" comunicate with the other nodes, it started to consume 100% of one core for each node.

I mean from 400% cpu capacity, rt_qmon ending consuming 300%.

No matter in what sequence the nodes are booted

Also if I start to stop the nodes the cpu resources of the OT remain in 300%.

I'm really worried about this, are there anything that I can do to solve this.?

Eduardo

Hi Claes,
I installed from scratch ubuntu 12.04 (this solved the cpu resources issue).

Returning to the "Fatal IO Error 11" failure.
I would like to mention that, I tried different alternatives without success.

First of all I want to comment, because I didn't mention before, that I am using OT's with dual monitor, extended desktop (HDMI-DVI).
Each graphics on each monitor, contain about 25 Ge pwr_valuelarge subgraph.

1 - Perform upgrade the nvidia card.
2 - I replaced the graphics card nvidia GE Force520 for another AMD Radeon HD5450 card, using the latest driver.
3 - I tried different versions of graphics drivers.
4 - I tried on one OT, using one single monitor configuration.
5 - I changed the Unity environment from Gnome Unity to Classic.
6 - I replaced the OS on one of the two OT's from Ubuntu 12.04 to Debian 7 and tested Gnome and KDE.
7- I Installed OT on a virtual machine.

In all the above points, after about a day, the failure repeats again (Top shows Xorg with 2,3%Cpu, and 11.8%Mem).

One thing, I disconnected the network cable on one OT Station and eventhough graphics are not being updated, the failure didn't repeat again.

Another test I did was to increase the following GeGraph parameters, SCANTIME, FastScantime, AnimationScantime from 0.5 to 1.5sec. The result of this modification, extended the occurrence of the failure from one day to three days.

The only way that the GeGraph works without any problems is when the graphics are open in the Process Station.

I think the problem come up when the Xorg (GeGraph) receive data from the ethernet conexion, no matter which OS or Desktop enviroment or the Graphic card I'm using.

Questions:
Do you know anyone with a similar configuration, that has this kind of issue?
Can I do another tests?
Could the PS has any misconfiguration?

Greatly appreciate any help.

Eduardo