Forum Replies Created
-
-
21st May 2007 at 11:59 pm #14655
Just a note:
Configuration parameters are stored in the dispatcher.conf, this includes every parameter you’re able to configure through the “Configure Dispatcher” icon in Muster explorer.
Every kind of different data is stored in the database.
21st May 2007 at 11:58 pm #14654There’re no command line switches in MExplorer. We keep your question as a suggestion for the future.
Regards.
21st May 2007 at 11:58 pm #14653We are aware of the reaction speed. A big difference in Muster 5 is that now data resides on a SQL db, even when it’s configured to work with the in-process sqlite. Just to explain that using the same system, it will be impossible to reach the speed of version 4.X that works in RAM and dumps periodically the queue status. Using an SQL database gives maximum file tollerance with the support of a journaling system.
The just-released version 5.2 address some speed issues, the network protocoll react faster and we optimized lot of the internal queries. You’ll find a huge improvement in the submission and the reinit of a job as well as the selection logic speed that were the critical data operations in the past.
Regards.
21st May 2007 at 11:55 pm #14652We confirm some uninstall issue. It’s always a good idea to manually uninstall the previous version, remove the Virtual Vertex folder and then install a fresh copy.
regards.
21st May 2007 at 11:54 pm #14651Exit code 211 always means an error reading the job’s files. this may have multiple reasons, the most common is related to permission problems for the service, in that case, you’ve to check the Muster user (or any custom user you supplied during the installation) and be sure it’s able to map drives and access network resource.
When it happens on a random basis, the most common source of the problem is the “LIMITED” amount of concurrent access you actually have on your file sharing host. I.E. WinXP Professional allows only 5 concurrent connections from external hosts.
It may be possible that your file server has such kind of limit.
Regards.
21st May 2007 at 11:51 pm #14650This is already configurable. You can simply open the template file installed in your dispatcher service (in the templates directory) and locate the last section of the template.
It contains a keyword list for the “Error” and “Warning” status as well as the exit codes.Regards.
16th April 2007 at 12:14 pm #14640Maya 6.5 has a bug in the commandline. the -rfb flag used by Muster is expected to be “-rbf”. If you plan to use version 6.5, you’ve to derive a new template from the default one and change the command line according.
5th April 2007 at 3:11 pm #14639This means that they are selected in a reverse order ? This could change the problem. Knowing if it’s working reversed or not working at all can help us in detecting the source of the issue.
30th March 2007 at 2:27 pm #14637We are not aware of such kind of problem even we didn’t test with Fedora Core 6. We’ll setup a test installation with this release to see if it’s a problem with this distro.
30th March 2007 at 2:26 pm #14636Even users are running Muster on Vista, we are aware of some problems using the network drives. We are working on it, full certified Vista support is scheduled to be available in version 5.2.
We’ll release version 5.12 before that version, so it may take still some time.
17th March 2007 at 8:20 pm #14632The broadcast presence function is required by the DISCOVERY function in the management console. If you’ve configured and discovered your hosts, there’s no need to keep it active.
By the way, it’s an UDP broadcast packet, so , there’s no way to route it elsewhere than your local network.
17th March 2007 at 8:19 pm #14631Are we talking about the server or the clients ?
17th March 2007 at 8:18 pm #14630If you get a red dot, there’s an error reported by a renderclient. Take particular attention, when you publish a new template, it has a new template ID, it means that you’ve to reconfigure the PATHS for that particular template on each render client.
17th March 2007 at 8:17 pm #14629I think Mike is talking about the selection priority of the nodes. I’ll take a look over this and see if there’s something that’s not working as intended.
17th March 2007 at 8:16 pm #14628Having some log dumps may help everyone in this forum to understand what’s going on.
-