Public Alpha Testnet Technical Update. This is the first in a series of technical updates for our Community Testers on Discord.

17 Aug 2023, 17:01
📣 Public Alpha Testnet Technical Update This is the first in a series of technical updates for our Community Testers on Discord 🎉 The #NuNet team hopes to provide a highly technical update to our testers each week to update them on what the Development team is working on 🔧 You can become one of our testers by Joining our Discord This Week's Summary 🧑‍💻 1️⃣ Work is ongoing to complete the changes to the logging mechanism for ML jobs requested via the SPD. This has been a fairly major piece of work as we have had to change the code and synchronize the changes across four components SPD, DMS, Oracle and CPD. As well as the logging, the telemetry has also been updated to use elastic search cluster, again this was a fairly big job as we implemented a new authentication mechanism for each DMS to update with own API key, this should increase the reliability. 2️⃣ Telemetry has moved from Stats DB to Elastic Search and Kibana. The old stats DB has been deprecated and removed. We have added new telemetry items and all actions on the DMS are now sent to elastic search including a heartbeat message so that currently active devices can be identified. This means that statistics can be almost real-time (in terms of number of active nodes on the network and currently utilized resources.) 3️⃣ Fixed Issues with SPD Web app reporting "Websocket already in a closing state" The comms between the web app and DMS were interrupted when trying to run a second job the webapp now keeps the socket open to be able to run multiple jobs back to back. 4️⃣ ML job logs moving from Gist to Logsbin Both the Oracle and DMS have had updates to use a new logging provider. We moved away from github gists that were originally implemented as a quick way to start testing. We are still finalizing the testing on this change that has to be coordinated between both services and is not backwards compatible with the old gist based system. 5️⃣ SPD and CPD changes to accommodate the new logging functionality Finalizing the changes to incorporate the logging changes above 6️⃣SPD Peer list update SPD page to show list of current peers is being fixed to properly report all DHT peers and their onboarded resources, this will allow anyone running the SPD to quickly see what compute resource is available to their machine. This is a very technical update so if you have questions please let us know & we can create additional #TechEd content to help our community understand these updates 🧠

Same news in other sources

1
NuNet
NuNetNTX #766
Telegram
17 Aug 2023, 17:01
📣 Public Alpha Testnet Technical Update This is the first in a series of technical updates for our Community Testers on Discord 🎉 The #NuNet team hopes to provide a highly technical update to our testers each week to update them on what the Development team is working on 🔧 You can become one of our testers by Joining our Discord - This Week's Summary 🧑‍💻 1️⃣ Work is ongoing to complete the changes to the logging mechanism for ML jobs requested via the SPD. This has been a fairly major piece of work as we have had to change the code and synchronize the changes across four components SPD, DMS, Oracle and CPD. As well as the logging, the telemetry has also been updated to use elastic search cluster, again this was a fairly big job as we implemented a new authentication mechanism for each DMS to update with own API key, this should increase the reliability. 2️⃣ Telemetry has moved from Stats DB to Elastic Search and Kibana. The old stats DB has been deprecated and removed. We have added new telemetry items and all actions on the DMS are now sent to elastic search including a heartbeat message so that currently active devices can be identified. This means that statistics can be almost real-time (in terms of number of active nodes on the network and currently utilized resources.) 3️⃣ Fixed Issues with SPD Web app reporting "Websocket already in a closing state" The comms between the web app and DMS were interrupted when trying to run a second job the webapp now keeps the socket open to be able to run multiple jobs back to back. 4️⃣ ML job logs moving from Gist to Logsbin Both the Oracle and DMS have had updates to use a new logging provider. We moved away from github gists that were originally implemented as a quick way to start testing. We are still finalizing the testing on this change that has to be coordinated between both services and is not backwards compatible with the old gist based system. 5️⃣ SPD and CPD changes to accommodate the new logging functionality Finalizing the changes to incorporate the logging changes above 6️⃣SPD Peer list update SPD page to show list of current peers is being fixed to properly report all DHT peers and their onboarded resources, this will allow anyone running the SPD to quickly see what compute resource is available to their machine. This is a very technical update so if you have questions please let us know & we can create additional #TechEd content to help our community understand these updates 🧠
Public Alpha Testnet Technical Update. This is the first in a series of technical updates for our Community Testers on Discord.
📣 Public Alpha Testnet Technical Update This is the first in a series of technical updates for our Community Testers on Discord 🎉 The #NuNet team hopes to provide a highly technical update to our testers each week to update them on what the Development team is working on 🔧 You can become one of our testers by Joining our Discord - https://discord.gg/f3Xf7DnE This Week's Summary 🧑‍💻 1️⃣ Work is ongoing to complete the changes to the logging mechanism for ML jobs requested via the SPD. This has been a fairly major piece of work as we have had to change the code and synchronize the changes across four components SPD, DMS, Oracle and CPD. As well as the logging, the telemetry has also been updated to use elastic search cluster, again this was a fairly big job as we implemented a new authentication mechanism for each DMS to update with own API key, this should increase the reliability. 2️⃣ Telemetry has moved from Stats DB to Elastic Search and Kibana. The old stats DB has been deprecated and removed. We have added new telemetry items and all actions on the DMS are now sent to elastic search including a heartbeat message so that currently active devices can be identified. This means that statistics can be almost real-time (in terms of number of active nodes on the network and currently utilized resources.) 3️⃣ Fixed Issues with SPD Web app reporting "Websocket already in a closing state" The comms between the web app and DMS were interrupted when trying to run a second job the webapp now keeps the socket open to be able to run multiple jobs back to back. 4️⃣ ML job logs moving from Gist to Logsbin Both the Oracle and DMS have had updates to use a new logging provider. We moved away from github gists that were originally implemented as a quick way to start testing. We are still finalizing the testing on this change that has to be coordinated between both services and is not backwards compatible with the old gist based system. 5️⃣ SPD and CPD changes to accommodate the new logging functionality Finalizing the changes to incorporate the logging changes above 6️⃣SPD Peer list update SPD page to show list of current peers is being fixed to properly report all DHT peers and their onboarded resources, this will allow anyone running the SPD to quickly see what compute resource is available to their machine. This is a very technical update so if you have questions please let us know & we can create additional #TechEd content to help our community understand these updates 🧠