Announcing the second beta release of SlimServer 6.0
For comparison's sake, here's my baseline:
5.4.1 RPM on Mandrake 10.1, AlienBBC v0.93 Transcoderless Beta, SlimScrobbler CVS. This combo has been up for a couple of weeks and is stable for all uses except Internet Radio (SlimScrobbler causes a crash while trying to get/submit info about what's being played). We don't use much Internet radio around here anyway.
Browsing 695 albums with 7596 songs by 704 artists
[EMAIL PROTECTED] jack]$ grep Vm /proc/`pidof slimserver`/status VmSize: 69464 kB VmLck: 0 kB VmRSS: 66416 kB VmData: 65208 kB VmStk: 268 kB VmExe: 12 kB VmLib: 3556 kB
I have two synchronized wired players, an SB1 and a SliMP3, currently playing music (Last Day Of June by the Finn Brothers if you must know :)
Sadly, my USB2.0 chipset has gone TU, so my RHEL3 VMWare used for the b1 test is inaccessible. Downloaded the Windows EXE of b2 instead and am running it on XPSP2. This laptop is a Centrino 1.4GHz with 1.5 GB of RAM, mounting the music from the Linux server via SMB over 802.11b.
Stop the service on Linux, start it on Windows (nice job on auto-configging the SP2 firewall, BTW). The little IE container pops up and I immediately switch to Fishbone, causing a JavaScript error. Line 72, top.document.getElementById(...) is null. Switch to Firefox, all is well.
Then I start looking at players and recall that the players are on a different subnet than the laptop. Briefly considering the options, fixing this looks like too much effort for a weekend morning so we'll be continuing this with SoftSqueeze.
Watching the database build with Process Explorer, it seems quite well-behaved. A steady 5-10% CPU, 50-52 MB of RAM. Handles are getting released. Refreshing Fishbone is interesting -- while the skin does load quickly, observe how long the extra couple of megs of RAM stay used.
http://www.monkeynoodle.org/tmp/fishbone-refresh-during-db-rebuild.PNG
At 30 minutes into rebuild, RAM usage has grown a couple of megs and CPU is spiking to 25% fairly often. Looks like it's blocking on LANMANRedirector a lot, which is what I'd expect -- I'm sure the db rebuild would be a lot faster if I was local to the music.
Tried out five consecutive Fishbone refreshes, it does release the extra RAM grab in due time, but I've driven peak working set to 59 MB.
45 minutes into the scan, found my first bug: switching from Browse Albums to Browse Artists, there's still an "All Albums" entry at the top of the list. Browse Years is a nice new feature, hopefully this will lead to fixing a long-standing wish to play an artist's albums chronologically (yes, it's in bugzilla).
http://www.monkeynoodle.org/tmp/all-albums.PNG
Is it my imagination, or is New Music using the filesystem dates? It's more-or-less accurately picking up what's new, even though this is the first time 6.0 has seen this music... coolness.
And database build is done at 50 minutes. 696 albums with 7617 songs by 704 artists, that's a bit off from 5.4.1's opinion. I don't see anything obvious... could be because I didn't do a wipe-and-rescan on 5.4.1 since last adding music, but it does the auto-rescan every night and seems to see everything. Anyway, CPU is 0%, private RAM is 57MB and peak working set is 64MB. Time for a Softsqueeze...
An error occurred while launching/running the application.
Title: Softsqueeze 2.0a14 Vendor: Richard Titmuss, et.al. Category: Download Error
Corrupted JAR file at http://localhost:9000/html/softsqueeze/SoftSqueeze.jar
as an applet: Java(TM) Plug-in: Version 1.4.2_06 Using JRE version 1.4.2_06 Java HotSpot(TM) Client VM User home directory = C:\Documents and Settings\jcoates
Proxy Configuration: Browser Proxy Configuration
---------------------------------------------------- c: clear console window f: finalize objects on finalization queue g: garbage collect h: display this help message l: dump classloader list m: print memory usage o: trigger logging p: reload proxy configuration q: hide console r: reload policy configuration s: dump system properties t: dump thread list v: dump thread stack x: clear classloader cache 0-5: set trace level to <n> ---------------------------------------------------- load: class com.slim.softsqueeze.Applet not found.
java.lang.ClassNotFoundException: com.slim.softsqueeze.Applet
at sun.applet.AppletClassLoader.findClass(Unknown Source)
at java.lang.ClassLoader.loadClass(Unknown Source)
at sun.applet.AppletClassLoader.loadClass(Unknown Source)
at java.lang.ClassLoader.loadClass(Unknown Source)
at sun.applet.AppletClassLoader.loadCode(Unknown Source)
at sun.applet.AppletPanel.createApplet(Unknown Source)
at sun.plugin.AppletViewer.createApplet(Unknown Source)
at sun.applet.AppletPanel.runLoader(Unknown Source)
at sun.applet.AppletPanel.run(Unknown Source)
at java.lang.Thread.run(Unknown Source)
Caused by: java.io.IOException: open HTTP connection failed.
at sun.applet.AppletClassLoader.getBytes(Unknown Source)
at sun.applet.AppletClassLoader.access$100(Unknown Source)
at sun.applet.AppletClassLoader$1.run(Unknown Source)
at java.security.AccessController.doPrivileged(Native Method)
... 10 more
and at that point, the kids are tired of playing with each other, so my testing is done. Maybe more later.
-- Jack at Monkeynoodle dot Org: It's a Scientific Venture... Riding the Emergency Third Rail Power Trip since 1996! _______________________________________________ Discuss mailing list [email protected] http://lists.slimdevices.com/lists/listinfo/discuss
