Spark load time is VERY long on most of our user’'s computers. Although several computers are outdated, it is slow on systems that should be able to handle it just fine (Win2k or WinXP Pro w/ all updates, minimal to no other startup items, Spark 2.0.3 or 2.0.4, P4 3GHz, 512-2GB of RAM).
Is there any way to speed it up? We need users to actually utilize Spark (with our Wildfire server), but if it is not set to load at boot, then they will be reluctant to use it. So if your answer is to not load it at boot, please give me some other options.
I know from experience that the JRE is SLOW on Windows in most instances. Can Spark use the MS Java virtual machine instead of the Sun Java setup?
Is there a configuration option I can tweak to speed it up?
Please help, as users are demanding their Yahoo Messenger back because Spark is so slow, but we need them off of non-closed IM networks for security and compliance.
Is there any way to speed it up? Yes there is. Give it more memory for startup. This can be done using spark.exe.vmoptions (see here) - you should set at least these values (I didn’'t check if these are still fine for Spark 2.x or if you need even higher values):
Can Spark use the MS Java virtual machine? As far as I know MS’‘s VM is outdated and MS recommends to uses Sun’'s JVM.
Java version 6 will help a lot with startup time and general performance. It’‘s currently in beta and will be released in time for the next major version of Spark. In the meantime, hopefully LG’'s suggestion will help. For users with less powerful computers, you could also resort to trying another XMPP client.
Well, still having speed issues. I will try out the settings mentioned in an above post. I did update to the latest official JRE release on computers that had the most problems. This didn’'t help any.
Which other XMPP clients exist that lack the ability to connect to public IM networks (AIM, Yahoo, MSN, etc) that are Windows clients, non-Java? I use GAIM at home and love it, but it would once again allow more intelligent users to access the public networks.
I have found the quickest way to speed up SPARK is the JRE RC6 suggestion above. Also, please rename your “jre” folder in Program Files\Spark to something such as “old_jre” and also set RC6 as the only JRE Runtime application through your Java CPL.
Just to let you guys in on some info, the next big version of Spark (2.5) will be using JRE 1.6 which is faster than lightening :). I’'m using it exclusively right now, and the speed improvements are dramatic.