| Devin Bougie wrote: |
|
Hi, All. We have been successfully using ELOG in a limited deployment for a couple years now. However, we are about to embark on a new project that could run for up to 10 years, and are wondering what sort of scalability we can expect from ELOG.
Are there any problems we can expect to run into as the number of entries grow? I see in a previous thread that "elog runs fine for a few 10000 entries. At 100000 entries it starts getting slow." Is this still the case, or have any improvements been made? What sort of problems would we expect to run into? Any examples of existing large deployments would be very useful.
|
The above made statement is not true any more. Mainly due to the large CERN experiments, some speed improvements have been made in late 2007. So elog runs fine at least up to 100000 entries. The startup time might be a bit slow, since it parses all entries there, but beyond the maybe 20s startup time, there is not a big difference when browsing entries. The only peoblem left is if you try to search some text through 100000 entries, this could be a bit slow. I have not tried anhything beyond 100000 entries, because this was not requested so far. If logooks become too big, the entreis could be split into several logbooks. Even if there are several logbooks with 100000 entries each, the access time should not be slower than if there would be one logbook with 100000 entries. |