14 | Join performance in ObjectdbSELECT count(e) FROM Meta e, ImageNet i, Page p WHERE e.idImage=i.id AND e.idPage=p.id; So simple ... of an Image and a Page instances instead of direct references. A more JPA conventional variant of your query would be: SELECT count(e) FROM Meta e, ImageNet i, Page p WHERE e.image=i AND e.page=p |
13 | How Should I Configure objectdb.conf to Obtain 256KB Disk IO Requests and Maximize Shared PersistenceManager Entity Cache? caching 1 MB (4 256KB sized pages). The goal is to have frequent PersistenceManager L1 cache hit ... above) appropriate to achieve this goal? CBE CBE Having larger database pages may have other effects (e.g. searching for an object in a larger page could take longer). So, as with most tuning |
13 | Iterating over large result set in memory was quite a challenge even with 3GB heap space. So I did paging using two loops. for (int page=0; page < resultCount; page+=pageSize) { Query q = em.createQuery("query"); for (Tuple t : q.setFirstResult(page).setMaxResults(pageSize).getResultList()) { // work |
12 | Inserted entities with strings as keys and indices needs more and more ram memory in comparing to primitive integers as keys and indices are sequential and therefore they are stored in sequence pages in the database. New objects with UID ... may require searching positions in the database and processing much more database pages ... is wasted in database when the storing objects sequentially, because every page be filled |
3 | Use temporary files to enable very large transactions results), currently the size of a transaction (i.e. the total size of database pages ... your needs. Particularly the unlimited size. The memory problem is with Page instances in memory ... of Page instances. support Support Regarding a new maximum for transaction size. In the long term |
3 | each 1-2 days objects on some objectdb level lock/block each other The two thread dumps show two different waiting situations. The first thread dump shows a page cache lock ... ="*" level="info" /> </general> <database> <size initial="256kb" resize="256kb" page ... application they may refresh the page if the response is too slow (again and again, as |
2 | Database access error , Doctor hanging on trying to repair, production shutdown at our biggest customer ------------------ [1] com.agile.hummingbird.ObjectProperty - Unexpected object count: 49830620 (actual 49830619) Page Content Errors ------------------- [1] Page #16748245 has a non first section entry (at entry 1 ... -------------------------- Group #1: Page#7880349 5:194359909 0/2 0+1954/2699 ... hgzwicker Hans-Georg Zwicker |
2 | Querying error - java.lang.ClassCastException: com.objectdb.o.STV small / large objects. ObjectDB splits large objects (> 2000 bytes) into pages. Somehow the stack trace indicates an unexpected state in which the same object is marked as small in one page and as large in another page. Maybe this is the result of switching between the two states when the object |
2 | [ObjectDB 2.6.9] Unexpected exception (Error 990): 142351618 (actual 142351616) Page Relation Errors -------------------- [1] Page #36099460 unexpected exception: null [2] Page #36099460 last key is null, parent page #43650945 next key is ('3.203.10.00.100000','95798112',2421822 8) [3] Page #36230484 unexpected exception: null [4] Page #36230484 |
2 | Server out of memory exceptions side was stable. Memory usage on the server side became higher mainly due to increasing page cache size. I see that the test uses 4 different databases (2 large and 2 small). The default size of each page cache ... a little more). Other caches (page caches of the 2 other databases, query results caches, etc.) consume |
