Problem with Umbraco 7 update with several Servers in NLB cluster
Hello dear Friends,
I'm looking for the help with Umbraco 7.1.4 and distributedCall. In nutshell - from 4 servers in NLB cluster after night update the last one (permanent) becomes not updated.
Details:
Soft: Windows 2012R2, MS SQL 2012, Umbraco 7.1.4
Architecture: 4 Server in NLB cluster and IIS Shared Conf settings + 1 Server as SQL and Shared Storage where site lives. All IIS's look in that share, except TEMP and LOGS directories.
Additinional Soft: my own code to update/add new nodes in Umbraco which works like a charm on standalone Umbraco.
Problem: Every night I use external script to call pages on main Umbraco and update nodes.(I think I can skip this part while I think it's not so important right now). As final result I get 3 out of 4 Umbraco sites updated and the last one (all the time the same one) with stale data. Only when I delete cache file manually, it returns actual data. I want to pay attention that all IIS's are the same cause of Shared Configuration and one place for the site.
Additional Conditions: For the sake of performance I decided to turn off cache&distributedCall in umbracoSettings.config before calling my code and set everythign back after with calling RePublishAll as the last action. (contentService.RePublishAll() returns True all the time).
Additional info: There exists Umbraco 4 on the same platform (with previous version of updating code) and I get ALL 4 server updated! So I dont' think it's platform issue, but I don't touch any settings with cache&destCalls on this Umbraco.
Question: So, what is wrong with 4-th server\Umbraco? Or maybe it's something wrong with the 1-3? I appreciate any suggestions and ideas! If you know how to turn on extra super loging or any other magic to helps me to understand the nature of problem - please, share it.
P.S. Obvious solution would be not to touch cache&distCalls but it lets me to get it done twice faster! And logically it sounds weird ... I'm going to test it and report results soon.
As I promissed here is the repord about "not to touch cache&distCalls" - as expected - everything is OK.
Every node(server) returns actual data after night update.
BUT it's taken more then 2 times longer as usual ... :( while there is not huge amount of data - I can live with it, but I hope you agree that it's odd to populate every node in cluster while I can do it with one and then just republish all at once.
P.S. Maybe I'm wrong :) but I just wonder about situation when in very close future we will have to deploy 8, 16 or even 32 hosts in NLB, how much time will it take?
Problem with Umbraco 7 update with several Servers in NLB cluster
Hello dear Friends,
I'm looking for the help with Umbraco 7.1.4 and distributedCall.
In nutshell - from 4 servers in NLB cluster after night update the last one (permanent) becomes not updated.
Details:
Soft: Windows 2012R2, MS SQL 2012, Umbraco 7.1.4
Architecture: 4 Server in NLB cluster and IIS Shared Conf settings + 1 Server as SQL and Shared Storage where site lives. All IIS's look in that share, except TEMP and LOGS directories.
Additinional Soft: my own code to update/add new nodes in Umbraco which works like a charm on standalone Umbraco.
Problem: Every night I use external script to call pages on main Umbraco and update nodes.(I think I can skip this part while I think it's not so important right now). As final result I get 3 out of 4 Umbraco sites updated and the last one (all the time the same one) with stale data. Only when I delete cache file manually, it returns actual data. I want to pay attention that all IIS's are the same cause of Shared Configuration and one place for the site.
Additional Conditions: For the sake of performance I decided to turn off cache&distributedCall in umbracoSettings.config before calling my code and set everythign back after with calling RePublishAll as the last action. (contentService.RePublishAll() returns True all the time).
Additional info: There exists Umbraco 4 on the same platform (with previous version of updating code) and I get ALL 4 server updated! So I dont' think it's platform issue, but I don't touch any settings with cache&destCalls on this Umbraco.
Question: So, what is wrong with 4-th server\Umbraco? Or maybe it's something wrong with the 1-3? I appreciate any suggestions and ideas! If you know how to turn on extra super loging or any other magic to helps me to understand the nature of problem - please, share it.
P.S. Obvious solution would be not to touch cache&distCalls but it lets me to get it done twice faster! And logically it sounds weird ... I'm going to test it and report results soon.
Thanks in advance. Looking forward for the help.
Hello Everyone,
As I promissed here is the repord about "not to touch cache&distCalls" - as expected - everything is OK.
Every node(server) returns actual data after night update.
BUT it's taken more then 2 times longer as usual ... :( while there is not huge amount of data - I can live with it, but I hope you agree that it's odd to populate every node in cluster while I can do it with one and then just republish all at once.
P.S. Maybe I'm wrong :) but I just wonder about situation when in very close future we will have to deploy 8, 16 or even 32 hosts in NLB, how much time will it take?
is working on a reply...