| Summary | |
|---|---|
| Article Number | 25824 |
| Environment | Orbix 6.x on all platforms |
| Question/Problem Description |
When running itadmin nsog list_members <Group Name> the members are listed in the first column and there are two other columns. It may not be clear exactly what the numbers in these columns mean. |
| Clarifying Information | An example of listing object groups looks like this: > itadmin nsog list_members StockFeedGroup
hello:RR_Member1 0.000 123456
hello:RR_Member2 0.000 1000000000
This is only the the case for object groups with active load balancing. This is not seen for object groups with round robin or random load balancing. The onus is on the service implement the object reference in this member to update the load as appropriate. This is not done automatically by Orbix. |
| Error Message | |
| Defect/Enhancement Number | |
| Cause | |
| Resolution | The second column is the load on the member and the third column is the timeout on the member. The timeout can be set to -1 to avoid a member being removed from the object group. The load and the timeout of the members can be updated and retrieved using the following methods on the ObjectGroup interface.
void update_member_load(
in MemberIdList ids,
in double curr_load
) raises (NoSuchMember);
double get_member_load(
in MemberId id
) raises (NoSuchMember);
void set_member_timeout(
in MemberIdList ids,
in long timeout_sec
) raises (NoSuchMember);
long get_member_timeout(
in MemberId id
) raises (NoSuchMember);
References to other documentation: Orbix Programmer's Guide, "Object Groups and Load Balancing". |
| Workaround | |
| Notes | |
| Attachment |
| Created date: | 12 October 2011 |
|---|---|
| Last Modified: | 13 February 2013 |
| Last Published: | 12 May 2012 |
| First Published date: | 19 October 2011 |
#Orbix
#KnowledgeDocs
