Latest Contributions by DavidKlein1

Like many organisations, we have an editing database and a publication database, and we use Esri 1way replication to keep the versioned feature classes in sync.  To complicate matters, we have geometric networks and many relationship classes as part of the replica.   We have long observed that some of the relationships do not make it across - that is feature x is related to feature y on our editing database, but they are not related on the publication database, aka the replica target.  We have never been able to reproduce this, and when investigated in the past, the discrepancy counts were low enough that we didn't think it warranted too much digging.   However, we found today that at least for one subtype of one feature class, the results are pretty disastrous.  There are approximately 60,000 records of this particular subtype, all of which should have a related feature.  On the editing database, about 600 do not have the required related feature.  These are likely data errors, but only affecting 1% of the data...we can manage that.  But on the publication database, roughly 30,000 of these features are lacking their related feature!  Half!  And I of course expect we will find some similar ratios as we expand our analysis. So my question for all you out there in GeoNet world - has anyone seen this before??  Any ideas as to what might be happening? ... View more

Hello, Large Australian electric utility here.  We are in the process of upgrading from ArcGIS/ArcFM 10.0 to 10.2.1.  A brief outline of the current 10.0 Server environment: 37 map services 7 SOM servers 8 additional SOC servers 1000 web GIS users, typically 250 concurrent We made the decision some months ago to move all of our map services to a single site with a single cluster.  This site was to include 12 servers, running Win 2k12 with 4 core and 8GB RAM each.  Our test environment has only 2 servers in it, and has been performing OK through most of the project.  We had no reason to believe that going into production with 12 would be anything but much faster.  Then last week, Esri-AU came to us, freshly returned to Oz from the Esri Developer Conference, and said it is recommended to not include greater than 4 servers in any given site; that the chattiness between the servers can degrade performance.  Apparently this issue is fixed at 10.3, but that is not an option for us.  We need to find a way to make it perform at 10.2. Now, to try to substantiate this recommendation, we have added more servers to that test environment, and done a series of load tests with 1,2,4, and 8 servers in the site.  We have thus far seen little evidence that going above 4 is any real problem...we have only run into performance issues where we did not have enough threads for a given service or 2.  (and we know we can configure around this) My question is this:  how have other people  - especially large organisations - elected to lay out your ArcGIS Server 10.1/10.2 environments?  I am interested in: number of services number of users number of servers (including core count and RAM) number of sites number of clusters per site Any wisdom and experience anyone can pass on would be much appreciated!! David ... View more