[260714] 케이스를 바꿨더니, 미뤄뒀던 일이 한꺼번에 풀렸다
두 번째 그래픽카드를 드디어 제대로 꽂았고, 그 김에 가장 중요한 종목 두 개가 몇 주째 데이터에서 빠지고 있던 것도 발견했다 며칠 전 케이스 안에서 자리가 안 맞아 미뤄뒀던 일을 오늘 마무리했고, 그 김에 꽤 오래 조용히 묻혀 있던 문제도 하나 찾아냈습니다. 임시로 돌리던 두 번째 카드를, 오늘 정식 케이스에 앉혔다 며칠 전 그래픽카드를 하나 더 들였을 때, 케이스 안에서 카드 두 장이 들어갈 자리와 전원 공급 장치 자리가 몇 센티미터 차이로 부딪혀서 정식으로 케이스에 다 넣지는 못하는 상태였습니다. 그래도 일요일부터는 임시로나마 두 장을 다 연결해서 실제로 돌리고 있었고, 오늘 전원 공급 장치를 별도 칸으로 분리한 케이스로 교체하면서 드디어 정식으로 제자리에 앉혔습니다. 전력을 적당히 제한해두고 부하…
The day began with a second graphics card finally being properly installed, leading to the discovery that two key stocks had been missing from data for several weeks. Earlier, the reporter had mentioned the importance of tracking the collection of news for these stocks; however, they had noticed that the extraction of company names from the data was failing for these two large, important companies.
While the backup data was already collected, fixing the bug that prevented proper name recognition allowed the reporter to quickly recover thousands of missing entries.
After installing the new graphics card, the reporter conducted a power supply test and found that both graphics cards were functioning properly, with stable temperatures and no errors. They rebooted the computer twice during the process, but the background services automatically recovered without issue.
While working on the graphics card installation, the reporter realized that the unusual fan noise they had suspected earlier was actually coming from the case's fans, not the graphics card. This was a relief, as they had mistakenly thought the noise was an indicator of a problem with the card.
The reporter also discovered that the most important stocks were consistently missing from the news database due to a name recognition bug. The names of these two major companies were being left blank in the database, preventing any articles mentioning them from being properly indexed. Fortunately, since the original data was already collected, the reporter was able to fix the bug and restore the missing information in a matter of minutes.
Additionally, the reporter found that the news scraping tool sometimes blocked access to certain pages due to a specific phrase appearing in the article. This was due to the tool trying to block pages that contained the same phrase used in an IT article. The reporter adjusted the tool to prioritize pages with actual article lists over those with matching phrases, preventing unnecessary blocks.
Other issues discovered included API request failures when sending too many simultaneous requests, particularly affecting major stocks towards the end of the day. Reducing the number of simultaneous requests resolved the issue. The reporter also adjusted the timing of the automatic analysis start and ended the data collection process after regular trading hours, using regular market closing prices instead of after-hours prices. This ensured a more accurate and complete dataset for analysis.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
