Disk space outage on multiple runners

This normally-infrequent problem has suddenly happened a lot in the last two days.

Chart of occurrences in last 30 days

https://grafana.gitlab.haskell.org/d/167r9v6nk/ci-spurious-failures?orgId=2&from=now-30d&to=now&var-types=no_space&refresh=15m

Table of affected runners

Admin URL	description
https://gitlab.haskell.org/admin/runners/349	aarch64-linux-2.zw3rk_aarch64-linux-2_1e7249a3b8da
https://gitlab.haskell.org/admin/runners/531	azure-ghc-ci-linux-1
https://gitlab.haskell.org/admin/runners/532	azure-ghc-ci-linux-2
https://gitlab.haskell.org/admin/runners/534	azure-ci-1__9d97fee6dfe8
https://gitlab.haskell.org/admin/runners/569	aarch64-linux-2.zw3rk_aarch64-linux-2_97f12874a58d
https://gitlab.haskell.org/admin/runners/594	aarch64-linux-2.zw3rk_aarch64-linux-2_97f12874a58d
https://gitlab.haskell.org/admin/runners/595	aarch64-linux-2.zw3rk_aarch64-linux-2_97f12874a58d
https://gitlab.haskell.org/admin/runners/596	aarch64-linux-2.zw3rk_aarch64-linux-2_97f12874a58d
https://gitlab.haskell.org/admin/runners/599	aarch64-linux-2.zw3rk_aarch64-linux-2_97f12874a58d
https://gitlab.haskell.org/admin/runners/600	aarch64-linux-2.zw3rk_aarch64-linux-2_97f12874a58d
https://gitlab.haskell.org/admin/runners/601	aarch64-linux-2.zw3rk_aarch64-linux-2_97f12874a58d
https://gitlab.haskell.org/admin/runners/602	aarch64-linux-2.zw3rk_aarch64-linux-2_97f12874a58d
https://gitlab.haskell.org/admin/runners/603	aarch64-linux-2.zw3rk_aarch64-linux-2_97f12874a58d
https://gitlab.haskell.org/admin/runners/604	aarch64-linux-2.zw3rk_aarch64-linux-2_97f12874a58d
https://gitlab.haskell.org/admin/runners/607	x86_64-linux-3.zw3rk_x86_64-linux-3_db038e4e1623
https://gitlab.haskell.org/admin/runners/608	x86_64-linux-2.zw3rk_x86_64-linux-2_d292a225d2eb
https://gitlab.haskell.org/admin/runners/609	x86_64-linux-6.zw3rk_x86_64-linux-6_db038e4e1623
https://gitlab.haskell.org/admin/runners/611	x86_64-linux-4.zw3rk_x86_64-linux-4_db038e4e1623
https://gitlab.haskell.org/admin/runners/612	x86_64-linux-5.zw3rk_x86_64-linux-5_db038e4e1623
https://gitlab.haskell.org/admin/runners/613	x86_64-linux-7.zw3rk_x86_64-linux-7_db038e4e1623
https://gitlab.haskell.org/admin/runners/614	aarch64-linux-2.zw3rk_aarch64-linux-2_379242e21d29
https://gitlab.haskell.org/admin/runners/627	aarch64-linux-2.zw3rk_aarch64-linux-2_379242e21d29
https://gitlab.haskell.org/admin/runners/632	maurer.smart-cactus.org_maurer_9a0f9dcfb3bc
https://gitlab.haskell.org/admin/runners/635	aarch64-linux-2.zw3rk_aarch64-linux-2_379242e21d29
https://gitlab.haskell.org/admin/runners/637	aarch64-linux_ghc-arm-5_f8e114d3fbe1
https://gitlab.haskell.org/admin/runners/639	x86_64-linux_ghc-ci-x86-1_2cf4fa5ac879
https://gitlab.haskell.org/admin/runners/641	x86_64-linux_ghc-ci-x86-2_2cf4fa5ac879

What's striking is that some of runners have identical descriptions, hinting at another problem! (Edit: #23095 (closed))