René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Speed up cold-cache offers","articleBody":"Requesting offers is quick if the offers cache is already populated but can be painfully slow if the cache is cold, e.g. 30s. Getting offers is one of the most common actions in dstack, so its performance affects the UX massively. We need to research how to reduce get_offers time for the slowest backends.\n\nOn dstack Sky, returning offers may take \u003e30s:\n\n```\ndstack apply  1.67s user 0.69s system 6% cpu 36.534 total\n```\n\nLocally, I get the following numbers for Sky backends:\n\n```\n[13:26:43] INFO     dstack._internal.server.services.backends:365 Requesting instance offers\n                    from backends: ['gcp', 'runpod', 'aws', 'azure', 'lambda', 'nebius',\n                    'cudo', 'verda']\n           DEBUG    dstack._internal.server.services.backends:404 Got offers from backend\n                    BackendType.RUNPOD in 0.551435s\n[13:26:45] DEBUG    dstack._internal.server.services.backends:404 Got offers from backend\n                    BackendType.LAMBDA in 1.839518s\n           DEBUG    dstack._internal.server.services.backends:404 Got offers from backend\n                    BackendType.VERDA in 1.817387s\n[13:26:47] DEBUG    dstack._internal.server.services.backends:404 Got offers from backend\n                    BackendType.CUDO in 3.355270s\n           DEBUG    dstack._internal.server.services.backends:404 Got offers from backend\n                    BackendType.NEBIUS in 4.055332s\n[13:26:48] DEBUG    dstack._internal.server.services.backends:404 Got offers from backend\n                    BackendType.GCP in 5.753253s\n[13:26:55] DEBUG    dstack._internal.server.services.backends:404 Got offers from backend\n                    BackendType.AWS in 12.271486s\n[13:26:57] DEBUG    dstack._internal.server.services.backends:404 Got offers from backend\n                    BackendType.AZURE in 14.012781s\n```\n\nFor some reason, Sky is slower than local for same backends.","author":{"url":"https://github.com/r4victor","@type":"Person","name":"r4victor"},"datePublished":"2026-01-20T08:39:33.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":1},"url":"https://github.com/3479/dstack/issues/3479"}