Umbraco: high memory usage for a small site
u

uros_sowa_01189

almost 2 years ago
First a few words about our setup: We are using the latest version of Umbraco v12.2.0 with Flexible Load Balancing on Docker containers running Ubuntu Linux. We are having currently one publisher (Backoffice) and one subscriber (front-end). We went with headless approach using Umbraco Content Delivery API. For media we are using S3. API content is served through Kestrel inside front-end docker container. Both containers (subscriber and publisher) are using the same instance of Sql Server. Installed packages in Umbraco: - Umbraco Workflow v12.1.1 - uSync v12.1.0 In the Backoffice we have currently two sites, with up to 15 sub-pages on each. We also used multilingual setup, each page has content for two languages. Every page consists of a few blocks on block grid. The size of the DB is around 190MB (so nothing huge in any case). Our issue (this can be reproduced on each docker containers restart): I just focues on the front-end/subscriber container here. The same can be also noticed on the Backoffice container (publisher). When front-end (subscriber) docker container starts, it uses around 160MB of memory looking at docker stats. Right after the first request is made to the front-end, memory jumps to around 280MB. About half a minute from the first request, memory jumps again to around 1,7GB. At the time of the big memory jump (from 280MB -> 1,7GB) I see bunch of messages in logs like: DeliveryApiContentIndex Write lucene doc id:{id}|{languageCode}, category:{category}, type:{type} So I would say the Examine index is the cause of the high memory consumption. I also verified Examine indexes in the publisher Backoffice (can't access front-end because Backoffice is disabled): - DeliveryApiContentIndex: DocumentCount=146, FieldCount=22 - ExternalIndex: DocumentCount=284, FieldCount=849 - InternalIndex: DocumentCount=297, FieldCount=1054 - MemberIndex: DocumentCount=0, FieldCount=0 Is maybe someone else also noticing high memory consumption?
Umbraco Upgrade Version 10.6.1 to 13.0.3 - SQL Timeouts
k

ktb_eqtrdev

over 1 year ago
My team and I have attempted an umbraco upgrade to the latest stable 13.x.x on a production environment (at the time of work sign-off). We maintain three deployment environments for our client: - An internal test and a prep-prod/UAT environment. - Production All of these umbraco applications are hosted in Azure in the form of Azure Web Apps. Pre-Prod and Prod are the closes architecturally speaking as each have two app web apps: 1) An Administrative instance for the umbraco backoffice 2) A public facing instance of the website An upgrade on all environments prior to production has been relatively painless - allowing Umbraco to applicate the necessary changes to facilitate the upgrade to 13.0.3. Unfortunately, the same can be said about production: The production database contains significantly more content that that of the non-prod environments and could be a contributing factor to the SQL timeout exceptions observed in the installation splash screen and in the trace logs *Troubleshooting * There was two key issues that we resolved together on pre-prod: Ensuring that the .NET stack version on the Azure Web app was updated to v8 (which I did) Adding and increasing the **DistributedLockingReadLockDefaultTimeOut **values in the appsettings. I also came across one example of the issues with an Umbraco 13 upgrade as described in this Umbraco forms thread. The suggested remediation steps didn’t resolve my issue. If you have any ideas for next steps, I’d be appreciative. https://cdn.discordapp.com/attachments/1219659453012639856/1219659453155250196/fe0d208e-9f4d-4ae2-99b5-28a59d8d9fa0.png?ex=660c1b61&is=65f9a661&hm=a1dbe9908a72a51961dcb7b998f104e523489b853719e20c37f0b78a8229f598&